Second Interdisciplinary Workshop on Reimagining Democracy

Last month, I convened the Second Interdisciplinary Workshop on Reimagining Democracy (IWORD 2023) at the Harvard Kennedy School Ash Center. As with IWORD 2022, the goal was to bring together a diverse set of thinkers and practitioners to talk about how democracy might be reimagined for the twenty-first century.

My thinking is very broad here. Modern democracy was invented in the mid-eighteenth century, using mid-eighteenth-century technology. Were democracy to be invented from scratch today, with today’s technologies, it would look very different. Representation would look different. Adjudication would look different. Resource allocation and reallocation would look different. Everything would look different, because we would have much more powerful technology to build on and no legacy systems to worry about.

Such speculation is not realistic, of course, but it’s still valuable. Everyone seems to be talking about ways to reform our existing systems. That’s critically important, but it’s also myopic. It represents a hill-climbing strategy of continuous improvements. We also need to think about discontinuous changes that you can’t easily get to from here; otherwise, we’ll be forever stuck at local maxima.

I wrote about the philosophy more in this essay about IWORD 2022. IWORD 2023 was equally fantastic, easily the most intellectually stimulating two days of my year. The event is like that; the format results in a firehose of interesting.

Summaries of all the talks are in the first set of comments below. (You can read a similar summary of IWORD 2022 here.) Thank you to the Ash Center and the Belfer Center at Harvard Kennedy School, and the Knight Foundation, for the funding to make this possible.

Next year, I hope to take the workshop out of Harvard and somewhere else. I would like it to live on for as long as it is valuable.

Now, I really want to explain the format in detail, because it works so well.

I used a workshop format I and others invented for another interdisciplinary workshop: Security and Human Behavior, or SHB. It’s a two-day event. Each day has four ninety-minute panels. Each panel has six speakers, each of whom presents for ten minutes. Then there are thirty minutes of questions and comments from the audience. Breaks and meals round out the day.

The workshop is limited to forty-eight attendees, which means that everyone is on a panel. This is important: every attendee is a speaker. And attendees commit to being there for the whole workshop; no giving your talk and then leaving. This makes for a very collaborative environment. The short presentations means that no one can get too deep into details or jargon. This is important for an interdisciplinary event. Everyone is interesting for ten minutes.

The final piece of the workshop is the social events. We have a night-before opening reception, a conference dinner after the first day, and a final closing reception after the second day. Good food is essential.

Honestly, it’s great but it’s also it’s exhausting. Everybody is interesting for ten minutes. There’s no down time to zone out or check email. And even though a shorter event would be easier to deal with, the numbers all fit together in a way that’s hard to change. A one-day event means only twenty-four attendees/speakers, and that’s not a critical mass. More people per panel doesn’t work. Not everyone speaking creates a speaker/audience hierarchy, which I want to avoid. And a three-day, slower-paced event is too long. I’ve thought about it long and hard; the format I’m using is optimal.

Posted on January 8, 2024 at 7:03 AM78 Comments

Comments

Bruce Schneier January 8, 2024 6:55 AM

Key insights—Present

Civic spaces and deliberation

  • Nick Couldry: Public discussion spaces have become controlled by technological platforms (especially social media). Design principles of these platforms are flawed and undemocratic; they optimize for scalability instead of social value. Resonance as a better principle than scalability.
  • Woodrow Hartzog: Powerful actors in our society (big tech) routinely and constitutionally betray our trust at the expense of democracy. Tech platforms, for instance, have large incentives for unethical behavior, with adversarial outcomes such as loss of privacy, degradation of mental well-being, and misinformation.
  • Francesca Tripodi: “Deep stories”—stories that feel emotionally appealing but not grounded in reality—have always been present. Propagandists have identified how to spread new deep stories (e.g., #stopthesteal) through search algorithms, keyword curation, and strategic signaling. The “IKEA effect” of misinformation increases trust in misinformation: “You trust it more if you build it yourself.” Individuals are encouraged to do this through “independent research,” engaging in “participatory disinformation.”
  • Nathan Schneider: Some basic things about human life are not available in online contexts, such as making small talk, getting to know someone before having hard conversations.
  • Bruce Schneier: Political preferences aren’t on the ground to be discovered, they are formed through democratic participation.
  • Deb Roy: There’s a trade-off between AI efficiency and the necessity of directly participating in democracy, even if less efficient: one person’s bias is another person’s judgment; and making sense of the world, even if imperfectly, is a democratic act.
  • Bailey Flanigan: Whose interests are accounted for in deliberation-based decisions? Partly influenced by how people internalize the interests of others. Internalization of others’ interests influenced by: underlying similarity / homophily, preexisting biases, existing propaganda/polarization/deep stories, design of information/deliberation, who is in the room, how AI is integrated, cultural context.
  • Aditi Juneja: Only 3.5% of Americans engaged with some type of political organizing; if politics is a practice, we are not practicing. Democracy could be about values and rights and not all about elections, but this requires we practice it.

Traditional media

  • Misha Glenny: Surge of far-right political groups, populist groups, and so on, and have become much more efficient than they used to be at coordinating an assault on democracy. Viktor Orban playbook: as soon as you get in, undermine the courts, the media, and the cultural institutions before you go back to the ballot box.
  • Spencer Overton: The presence of a liar’s dividend (“politicians profit from an informational environment saturated with misinformation”) as deeply damaging for public trust.

Voting and elections

  • David Perry: Voting is an inherently normal human behavior, it has been present throughout millennia, even in medieval times, where there were a range of secret, clever voting systems intended to prevent voting blocs — but also often to keep power in place.
  • Nick Chedli Carter: The practice of US politics in the 21st century is increasingly online, top-down and consolidated, dehumanizing, commercially-oriented, polarizing, and dominated by the presidential election cycle. Voting matters – parties matter – but the systems and structures need to be improved. There is a better way!
  • Kathryn Peters: Legislative elections are plagued by extremism, gridlock, unequal representation, and influence, plus the fact that interest in running/skill in campaigning are not good predictors of governing ability.

Governance, institutions, and law

  • Ulrich Wagrandl: Liberal democracy is the symbiosis of two different principles: liberalism (autonomy of individual) and democracy (popular sovereignty). The attack that the democrat part is staging on the liberal part is not wholly unjustified. Liberal democracy is a limited democracy; it is constrained by a range of mechanisms, all of which are counter-majoritarian (e.g., Constitution, judicial review, institutions, including proliferation of independent agencies, and civil society like the press and media).
  • Sam Gill: Modern liberal institutions struggle to address problems that are temporally broad, spatially broad, and involve problems of domination. They also fall prey to both rational and liberal flaws. See, for example, child protection (an administrative system, whose liberal flaw is racism, and whose rational flaw is high modernism). One of the key threats to liberal institutions and trust in them is the attack on normative knowledge. From the illiberal political side, this manifests through saying that science is a political speech act. From the technology industry, this manifests through saying that authoritative knowledge is an impediment to innovation and self-actualization.
  • Joshua Fairfield: The common law is a map of meaning; this case is like that case, because humans have thought there is a link. That heat map of meaning is being replaced through AI. Courts are beginning to use AI, which will replace the heat map—allows cases to be supplied that say what “people mean”—become put into court opinions, standardized, and become a new network of meaning.
  • Marci Harris: Congress has pacing problems: external (fails to keep pace with emerging innovation), inter-branch (lags executive branch), internal (doesn’t employ modern practices and technologies).

Collective imaginaries and culture

  • Glen Weyl: Often narratives of technological inevitability (either a Neal Stephenson narrative, with information leading us into a virtual world, or a narrative of AI taking over everything), which should be questioned.
  • Robert Jordan Hall: It is by no means obvious that the epistemology of the west is true or obligatory.

Capitalism and work

  • Archon Fung: How should democracy regulate capitalism? Currently, the dominant view is that national popular sovereignty (i.e., big democratic state) regulates big capitalism. However, this often fails through regulatory capture by capitalists, leading to political inequality.

Bruce Schneier January 8, 2024 6:56 AM

Key insights—Future

Civic spaces and deliberation

  • Build the conversational muscle to have hard conversations, and relentlessly accept the possibility of being wrong.
  • An alternate definition of polarization, on the possibility for debate rather than the strength of the views.
  • Deb Roy: Leveraging ancient wisdoms (facilitated deep dialogue and deep listening) and digital technologies: hearing the humanity of others through smaller circles of trust.
  • Hahrie Han: Drawing from the growth of megachurches to identify drivers of civic engagement: the smallest reproducible unit (4–5 people) is a crucial component. Politics should be a practice, instead of a spectacle: something in which each citizen engages. Should start with dyads, triads, allowing participants to focus on interests, identities, preferences, to be more relational, negotiation-based, experiential, about interpretation and commitment.
  • Nick Couldry: New design principles are required for public infrastructures and discussions: smallest scale needed, spaces in between for discussion, maximizing variety and experimentation, building trust within communities, interconnection between platforms.
  • Eli Pariser: Digital spaces that have three goals: what is happening (sharing and seeing experiences), what we think about it (sharing and seeing opinions), and what we could do about it (sharing and seeing ideas and action steps), enabled through greater representation of views, and moderation of extreme viewpoints.
  • Wes Chow: Challenge of conversational digital tools is to replicate rather than flatten the context inherent to real-life discussions, notably intimacy gradients.
  • Jon Evans: Wide adoption is required to recreate democracy, and this needs to start on a small scale, with democratic testbeds. Traditional testbeds include civic groups, fan clubs, churches, grassroots communities… New testbeds include decentralized autonomous organizations, CityDAOs (for particular cities), “network state” movements, online spaces (Discord, subreddits).
  • Kathryn Peters: Forming citizen’s assemblies without waiting for permission, and deliberating about complex topics, coming to conclusions, and moving beyond the traditional advisory role of assemblies.
  • Dimitri Courant: Citizen’s assemblies can come in many different typologies: time (exceptional, occasional, event-based, regular, permanent), members (citizenry, elected officials, randomly selected officials), powers (agenda-setting, ratification). A range of corresponding models are possible, of which three are pure models in which an actor holds both powers (tamed consultation, radical democracy, representative klerocracy), and a range of hybrid models. Institutional design is therefore crucial, both to determine just distribution of power, and to ensure assemblies are resilient.
  • Dimitra Dimitrakopoulou: Citizen’s assemblies in small groups is crucial for learning, notably through sharing experiences instead of opinions and enhancing engagement and listening—and technology can help capture and make sense of these conversations. Using technology for citizen’s assemblies allows participants to consider wider diversity of citizen perspectives, ease burden of deliberative process (subdivide-synthesize), and make the process more transparent and accountable.
  • Helene Landemore: Six ways to use AI to augment democracy, particularly in citizen assemblies: selection and distribution of participants to ensure representativeness in citizen assemblies; facilitation; translation; fact-checking; clustering and organizing arguments; deliberation output summaries.

Traditional media

  • Francisco Hernández-Cáceres: Blending science and entertainment to reduce polarization, through game shows bringing together contestants from various backgrounds (notably political) and get them to work together.

Emerging technologies (including AI)

  • Ruthanna Emrys: Changing technological priorities from engagement to clear collective and individual thinking, democracy, individual customization (e.g., neurotypes, social styles, personal preferences).
  • Glen Weyl: Towards plurality: technologies to facilitate collaboration across differences. Inherent tension between diversity and collaboration (hard to work deeply with people with very different viewpoints), and plurality aims to push back this trade-off.
  • Riley Wong: Data collectives (e.g., data coops, trusts, coalitions, unions, and commons) can also inform practices around governance, membership, access, and usage—the types of collective can also imply certain sorts of governance.
  • Riley Wong: Collective access, usability, and responsibility over ownership is ultimately a stepping stone towards stewardship, consent, and agency over control.
  • Bruce Schneier: Ten ways in which AI could influence democracy, both for good and bad: educator, sensemaker, moderator, lawmaker, political strategist, producing arguments, cheap reasoning operator, law enforcement, propagandists, political proxy.
  • Deb Roy: Using AI as a tool for sensemaking (and potentially listening, deliberation, mediation), but always keeping the human in the loop in a supervisory role.
  • Tantum Collins: Machine learning for normative representations can allow for a movement beyond rules and standards, by capturing tacit knowledge. Has negative implications (e.g., dealing with bias, input at multiple levels, informing citizens of more complex representations)—depends on what is meant by interpretability (e.g., tax code is interpretable but most people don’t engage with it).
  • Iason Gabriel: Given the inherent differences in values between communities, two solutions for a beneficial AI are possible: (1) global exercise of an “ideal spectator,” where regulations and safety standards are defined, and the material benefits of AI are regulated to benefit all evenly, and (2) democratization of AI, where citizen assemblies and annotators represent various communities and their respective values.
  • Aviv Ovadya: It would be challenging to regulate AI at the nation-state level, given possible externalities. An alternative lies in AI/platform democracy: representative, informed, and, where appropriate, compensated participants, making binding collective decisions about policies or values for AI systems across national boundaries. Four key components to AI/platform democracy: transnational representative deliberation; AI augmentation of processes; democracy as a service, through third party organizations running democratic or plural processes; AI/platform democracies using these technologies.

Voting and elections

  • Danielle Allen: People thrive when they have a bedrock of empowerment: need to move to a democracy that is responsive, where people are seen. Necessary to reconceive representation (e.g., digital civic infrastructure), finding concrete use cases, finding the people with the capacities to implement them and scale them up.
  • Jon Evans: New forms of voting can be tested in democratic testbeds: quadratic voting, direct voting, liquid voting, mandatory liquid voting, participatory budgeting… As well as new kinds of constituencies, such as optional online constituencies: people choose to vote either in their geographic district or as a part of a self-sufficient community.
  • Paul Golz: AI could be used to lead social choice processes at scale, as highlighted by an experiment with chatbots (OpenAI grant) leading to 75% of participants feeling perfectly represented by statements identified, and 18% mostly represented.

Governance, institutions and law

  • Rob Ricigliano: Government and governance are not co-equal concepts: governance is much broader! To be healthy, governance needs an ecosystem of five overlapping, mutually supportive process: being together (building connection and complex knowing of each other), working together, weaving together (understanding and connecting our identities), caring together, deciding together.
  • Danielle Allen: In order to experience autonomy you need to have autonomy across the whole scale of your experience, and since our collective action problems come on different scales, you don’t just need non-institutional structures—you do need some connection between levels, both civil society and institutions.
  • Nathan Schneider: Towards “modular politics” (e.g., at Metagov), a mode of doing governance tooling, with interoperability between diverse governance forms, and self-governance rather than top-down control.
  • Robert Jordan Hall: Institutions need to be resilient to changes in complexity (in quality, scope), which is not feasible through complicated systems (enumerating rules, states of human behaviors as the French civil code has tried to do).
  • Hahrie Han: We need a model based on both collective deliberation and individual negotiations. We need people who are capable of negotiation on behalf of others, who are masters of their politics but not their ideology.
  • Niclas Boehmer: Inherent trade-off with automated decision-making systems (ADMs) between transparency and accuracy of decision-making. Developers should put explainability at the start, not the end of development of ADMs. Society (notably social scientists) should identify acceptable trade-offs (can be domain specific), to drive greater legitimacy for these tools.
  • Nils Gilman: Imagining a more positive view of segmentary citizenship, such as for neighboring countries, for greater political inclusion and flexibility, notably in the context of climate migration.
  • Helene Landemore: Nobody can be excused by ignorance of the law; this is a basis of law. We couldn’t do this if we don’t even understand the law, for example, if we replace the constitution with a piece of code.
  • Suzette Brooks Masters: Incorporating future generations in decision-making, through intergenerational fairness assessment tool (as in Portugal), roleplaying future generations (in Japan), building a capacity for strategic foresight, new future-oriented governance roles (e.g., Wales future generation commissioner).

Collective imaginaries and culture

  • Sorcha Brophy: Imagination very structured by current realities. For example, in a scenario where students were asked to imagine future healthcare, they immediately recreated insurance, and then created equity committees.
  • Aditi Juneja: The narrative environment we’re in shapes what we think is possible. Hollywood destroyed Americans’ trust in communities, governments (e.g., House of Cards) hence, we need new stories. For drastic change to occur, both the narrative and political environments need to be aligned.
  • Eugene Fischer: Fiction as a space to imagine alternative democratic institutions, for instance a direct democracy where the problem of educating the electorate is solved by the outcome of elections themselves: binding but frequently reiterated public referenda allow collective reaction to the results of past decisions. Individual voting power could be based on aggregated histories of past votes, to incentivize prosocial forms of electoral engagement and limit the influence of voting blocs.
  • Suzette Brooks Masters: The future can be a powerful depolarizing tool, as there are often shared aspirations for the future, with more common ground than in the present. Using speculative fiction can unite people in a way that doesn’t have to be jargony, that is highly accessible and democratic.
  • Ruthanna Emrys: Using science fiction as an alternative approach to deliberative speculation: question assumptions about future development, assumptions about how knowledge is created, to consciously imagine our descendants (genetic, social, intellectual, philosophical) living in our imagined futures.
  • Claudia Chwalisz: Moving beyond the current democratic paradigm, making it more human, maybe through a rights-based approach, reconnecting humans to the natural world, coincidentally reconnecting humans to one another.

Capitalism and work

  • Victoria Goddard: UBI is promising because it can provide the space and time in people’s lives to enable them to participate in activities beyond mere survival—notably, to engage in civic participation.
  • Helene Landemore: Infusing more democratic processes in corporate governance, there are two main alternatives: technocratic oversight boards (e.g., Meta, but unclear whether truly independent and representative) or citizen assembly representative of humanity providing direct inputs, faster than regulatory processes.
  • Archon Fung: Principle of affected interests: an individual should be able to influence an organization if and only if that organization makes decisions that regularly or deeply influence that individual’s important interests.
  • Woodrow Hartzog: Platforms must be held accountable to those they serve: because of the lopsided relationship between the two, platforms must have a duty of loyalty, care, and confidentiality.
  • Glen Weyl: Threat of socialization can push companies towards good social responsibility. Lots of policies could have encouraged this socially responsible direction—important to take action, both proactive and regulatory, that makes social responsibility possible.

Bruce Schneier January 8, 2024 6:58 AM

Panel #1: Conversation

Lecture

Nick Couldry: Professor of Media, Communications and Social Theory, London School of Economics

  • Book on the fundamental error that humanity made on spatial design
    • Allowing large corporations to design the space of the world
    • The space of almost all possible spaces for social interaction and democratic practice
  • Allowed them to design shadow spaces (platforms and apps) with two properties
    • They can be controlled, indeed exploited, by those corporations for their own profit
    • They bolt onto our actual world under conditions those corporations control
    • Condition that we use them because everyone does
  • Three fundamental design mistakes
    • Enabling a space of spaces on unlimited scale with unlimited feedback potential
    • Not thinking about the spaces in between our platform spaces
    • Prioritizing scalability, not social/political design
      • This is orthogonal to democracy as we understand it
        • Aristotelian idea
        • Hobbesian idea of social contract
      • Neither saw democracy as unfolding on a global setting
  • We can’t unimagine platforms or the Internet, but we need to rebuild in ways less risky to democracy
    • Think about platform space in a different way
      • Secure the spaces in-between, enabling friction, not seamless flow
        • Limit the chances of feedback loops
      • Trust more in the specific types of discussion/encounter that this protects
      • Values and purposes align more with specific community
        • Trusting more in subsidiarity
      • Give a larger design role to existing communities and to subsidy government/civil society
        • Not as hosts but as general sources of subsidy
        • What Zuckerman has called an intentional public infrastructure
  • Change the guiding principle: instead of scalability, choose resonance
    • Even if you don’t agree with others, you have a shared perception of problems you’re answering, social assumptions
    • Resonance occurs in physics when there’s the same frequency between various objects…
      • Interaction between the sound source and the properties of the object itself; no need for an external input to this
    • Today’s social media landscape seems to be built against resonance
  • Preventing AI from becoming “McKinsey,” using Ted Chaing’s notion
    • Why only see it in a framework driven by profit
    • Think in the same way about platforms
      • Be oriented by principles of what’s good for life/democracy → very different priorities
  • Design principles
    • Build to the smallest scale needed
    • Design the spaces in between
    • Maximize variety and experimentation
    • Trust communities
    • Don’t maximize time on any one platform (instead, maximize interconnections, including between online and offline)
  • Then we have a chance of protecting the commons
  • Why shouldn’t populations displaced by climate change get a voice in the decision about where they will be relocated to?

Francisco Hernández-Cáceres: Founder and CEO, Country Roads Studios

  • Can media/entertainment reduce polarization?
  • Personal significance
    • Born in El Salvador, in the middle of a civil war
    • Experienced a democracy in peril from the moment he was born
    • Was in a very conservative society
    • Sneaked and watched Will and Grace, other shows that inspired him, which champion diversity
  • Moved here and worked for a range of US media companies
    • Won three Emmys and became a US citizen
  • Started seeing the same level of polarization as he left behind
  • Went to Harvard Kennedy School because finds that current media does not try to solve this problem
  • Created Country Roads Studio; the mission of the studio is to produce films, documentaries, television shows, and online games that aim to reduce tribalism
  • Threefold approach
    • Every project is based on science and entertainment
      • Taking research from Harvard (Julia Minson, Joshua Greene) and Stanford with a creative lens
    • High impact
    • Non-partisan
      • Met with a range of stakeholders from different backgrounds
  • Projects in development
    • Bridge or Sink: game show
      • Jeopardy meets American ninja warrior
      • Two participants who are from extreme ends of the political spectrum
      • Either bridge or sink together
        • Get a question right and part of the bridge gets built, get a question wrong and you could fall into the water
      • Contestants either win or work together
      • Changing the narrative because if America fails, they fail together
      • Three different kinds of trivia: from benign to complementary to politically biased
        • With the third, trying to debunk myths from both sides of the political spectrum
      • Using this for academic research, going to show the pilot to people and see the impact on polarization
      • Building an app to enable people to chat with each other and bridge the gap
    • The Great American: documentary + reality series
      • Squid Game meets The Great British Bake Off

Eli Pariser: Digital Activist and Co-Founder of New Public

  • Public conversation is a particular kind of design problem
    • We can and must do it better
  • New_Public: accelerates the creation of thriving digital public spaces, along the lines of what Nick described
  • Large-scale conversation is: critical for democratic societies, different from 1:1 and small group conversations, totally broken online, suited for PSM’s role
  • From toxic spaces to safe ones
    • We can reverse the cycle of most online spaces
    • Spiral of silence: Combative people crowd out reasonable ones
      • A small or combative group of people feels entitled or comfortable to post → combat dissuades others from joining
    • But if we can welcome them in, spaces become safer
      • A larger and more diverse group of people feels comfortable and welcome
      • Extreme views and behaviors are marginalized
      • Not necessarily all people who take part, but views are representative
  • What does a public conversation aim to do?
    • What is happening?
      • Share and see experiences
    • What do we think about it?
      • Share and see opinions
    • What could we do about it?
      • Share and see ideas and action steps
    • Need people to recognize their experiences and views in the conversation
    • Opportunities for public media organizations (private rule enforcer)

Deb Roy: MIT Center for Constructive Communication

  • Research, prototypes, pilots scaled deployments
  • Collaboration between MIT Center for Constructive Communication and Cortico
  • Cortico team
  • Have spent years studying social media
  • Everyone has a bullhorn, social fragmentation, most extreme points of view dominate
  • Hearing the humanity of others is necessary for democracy to function
  • Ancient wisdoms crossed with digital technology
  • Break a large group into smaller circles of trust, share experiences
    • Work with prompts to share experiences instead of facts. This strengthens social fabric more than mutual respect
  • Enabling listening at a distance
  • Larger groups, thousands: sensemaking, identifying themes, then sharing them back
  • Decision markers disconnected from communities and the communities themselves are disconnected
    • Need a commitment to listen and respond, and a commitment to participate
  • Conversation: in person or online through Zoom
  • Sharing and decision-making through voice portals
  • Example in Madison police department
    • Community dialogue → sensemaking → leadership selection
    • Maine: youth-led conversations
  • 100 partner organizations across the US
    • Local government
    • Youth-centered orgs
    • Emerging areas
  • Social dialogue network: facilitator, organizer, sensemaker, curator, conversation designer
    • Opposed to social media network for vitality; this is architecture that starts with resonance, not shouting
  • How to scale participation to millions
  • How to use AI to scale it
  • Fora: social dialogue network
    • Being able to capture the voices of real life conversations
    • Voice moving at the speed of trust
  • Using AI as sensemaking
    • Codemaking, AI generated codebook; AI tagging, thematic summaries
    • Interpretation depends on human perspective!
    • Social dialogues network, AI assistive tools, sensemaking and publishing, data privacy management, ancient wisdoms
    • Keeping the human in the loop—steer, inspect, evaluate at each step
  • Working on listening, deliberation (Democracy Next), mediation

Nathan Schneider: University of Colorado Boulder

  • Democracy on, not just around, the Internet
  • Cancel culture: absence of deliberative potential, deliberative capacity
  • Being a facilitator of in-person spaces, you understand that online spaces do need better facilitation
  • Basic things about human life are not available in online contexts
  • Design pattern of “implicit feudalism”: there is really just one way that power flows online
  • One person has all the power online; digital equivalent of censorship and exile
  • This is constraining democracy in daily life and at the largest scales
    • Tocqueville: everyday practices of democracy appear to be integral to our attempt at making democracy work at the largest scales
    • Sysop and politician are becoming blended together
  • We are crowding out the richness of historical governance legacies
    • “Governance archeology”: mapping pre-digital governance practices to rethink how we do digital governance
    • We are unable to access the legacies we could be drawing on, regarding how power flows in digital spaces
  • We could instead design systems for diverse forms of self-governance
    • Just trying to do basic democratic practice required breaking all the defaults, just to replicate the basic governance occurring at a smaller scale
    • Metagov: “Modular politics”: mode of doing governance tooling; interoperability…
  • What if policy encouraged self-governance rather than top-down control?
    • Moderation accountable to users
    • Worker organizing for all gig work
    • Community-owned infrastructure
    • Public options for AI
  • Book: Governable Spaces: Democratic Design for Online Life, University of California Press

Questions

  • Nils
    • Two assumptions:
      • That people do want to engage in discussions
      • That systems were not built to shut up specific views
    • What to do about people who don’t want to engage in discussions and do want to engage in canceling? Those people think they’re heroes in their own stories
    • Eli
      • Influence operations question that’s outside of the scope
      • Majority of bad behavior in spaces
        • People are situationally dependent actors most of the time
        • Quality of civic indifference: I’m not going to get involved in this conflict
          • A useful quality for pluralism
        • He does believe that you can build spaces such that it decreases the odds of people getting into “war mode”
    • Nathan
      • Lack of due process and governance mechanisms is what emboldens intolerant types
      • Legal system at least requires specificity of complaint and punishment, whereas in the online space there is no option except the most extreme type of removal
      • “The Tyranny of Structurelessness” by Jo Freeman
      • If we had these types of mechanisms, we wouldn’t need to reduce polarization
      • Moments of polarization are when change can happen
        • Leveraging these moments for change rather than empowering extremes
    • Deb
      • We have always relied on implicit friction to settle polarization, and we have now connected everyone frictionlessly
      • When we have connected spaces with no spaces in-between, can’t effectively bridge differences
  • How to define polarization less on strength of the views and more on the possibility for debate
  • Nick
    • We need polarization, otherwise we wouldn’t have politics
  • Francisco
    • Ronald Heifetz: how to solve problems
      • Conflict is good, we need productive conflict
      • You need people to be able to disagree in a compelling, respectful way
  • Francisco: Networks
    • Living in Ohio, trying to understand different context. Friendliness of people of Youngstown, Ohio
      • Our systems are designed for small-scale conflicts
    • At the core, most humans are good; but that doesn’t mean there aren’t nefarious people out there
    • His own parents are not going to his wedding next year, but they’re still good humans; they just don’t know how to love unconditionally
  • Other Emmy award winner in the room; networks?
    • How to find a network that is willing to invest in such a show
    • Haven’t found one so far
  • Judith:
    • In public conversation we’re looking for ideas not individuals? Pushing back against this
    • There’s an excess of efficiency online
    • When people understand words, it’s often in the context of the individual, of who is saying it
    • All of what seems wasteful, like small talk and so on is actually crucial; it helps us make sense of the people
  • Bad actors: very few people?
    • Our systems have been designed for small-scale conflicts
    • How to think about consensus, not just deliberation
      • Complexifying, bringing in more interdisciplinary approach
  • Bruce Schneier January 8, 2024 6:59 AM

    Panel #2: Past and future

    Lecture

    Eugene Fischer: Science fiction writer

    • Ionia, experimental polity ninety years in the future when there are large regions made uninhabitable by radiation
    • Ionia was started in one of these regions using gene therapy (cells can repair DNA damage from radiation)­
    • Becoming Ionian was a practically permanent decision; you can’t leave it, since you’re radioactive
    • Ecologists: saw government as a biological adaptation, a component of humanity’s extended phenotype
      • Trade-off between efficiency and resilience
      • Made an intentionally inefficient experiment in direct democracy
      • Electorate educated by the results of frequent but relatively inconsequential elections
    • Every Tuesday in Ionia was election day (ballot propositions, electronic voting, homomorphically encrypted ballots)
    • Ionian elections were based on aggregated voting histories rather than individual votes
      • Because minority positions were important as well
      • Minority threshold was 20%
      • One person four votes, your vote is only used if you win the proposition
      • After four weeks voting on the winning side, your votes count only once
      • Inherent instability balanced by continuous feedback
      • A system where power changes incentivized people to find broadly acceptable, and therefore lasting, solutions
      • Individuals incentivized to not vote on issues they don’t care about, to maximize their own future voting power
      • Ballot items added by legislators who are selected by sortition
      • The norm, when Ionia thrived, was for the ballot to remain stable
      • Weekly elections a collective ritual of affirmation
    • Children of Ionia did not always grow up to be as happy with isolation from the rest of the world as their parents
    • Loss of barrier to entry: after sixty years the region was decontaminated and Ionia was forcibly re-integrated into broader society

    Victoria Goddard: Fantasy novelist

    • UBI is argued for in one of her books—the main character cared about it, which made me care about it as well—more common way for people to become invested in particular issues
    • She’s not a polemical or didactic writer
      • And she ended up caring for UBI
    • Two key authors
      • Dante
      • Philosophical writer in the Roman empire
    • She’s not trying to write a book of politics, but as a civic animal, her book ends up being a part of politics
    • Where she lives, small community but considered a full commune
      • Not everyone actually takes part
      • How to integrate into this community
      • She became a representative in her community
      • Core part of democracy: random grouping of people
    • Small scale community work
    • People are often really busy and don’t have the time to get involved in these activities
      • Need to give people the space in their lives to do so, have the opportunity to care about things that are not basic survival! UBI achieves this

    Robert Jordan Hall

    • Causality of time, if we assume causality: he is engaging in orality
    • If reverse causality: holographic imagination
    • Relationship between things like technology and media and sociotechnical structure we live in, and psychological structure?
    • 500 years ago, invention of printing press
      • Contributed to Age of Enlightenment
      • We are in the very tail end of this
      • We are witnessing a level of technological innovation at a similar amplitude and likely much faster rate, and much more globalized, than what gave rise to the Enlightenment
      • We should imagine that the level of disruption to our institutions, ways of thinking, will be much greater as well
    • Design principles strong enough to traverse change
    • Complexity—observation of robustness of long-standing traditions (subtly, nuance, intimacy) over long period of time to build things that have a continuity and connectedness that can’t be modeled perfectly
      • Connectivity
      • Complexity evolves: can change in quality, in scope
      • Complicated systems have a tough time adapting to it
      • Institutions that we have are incapable of dealing with the situation and the scale of change happening
      • We’ll have to engage with the future, imaginaries like science fiction
    • Complicatedness: what can be expressed in semantic states
      • French civil code: trying to enumerate the state of human behaviors, and has failed to do so
      • Can only be updated by means of applied reason
      • Underlying notion that there is a “we” that can do things
    • Our assumed and inherited “we”s are no longer relevant
      • Polity, who
    • War
      • Mostly defined by kinetics
      • Some are engaging in the optimization of their local niche
      • Some are engaging in war
    • Diversity of epistemological groups
      • It is by no means obvious that the epistemology of the West is true or obligatory
    • Having the courage to recognize that the reality we are living in is coterminous with science fiction is fundamental

    Suzette Brooks Masters: Democracy Funders Network

    • Cross-ideological network of donors
    • Used to be highly pessimist but started seeing stories of good things happening
    • Better futures project at DFN: What animated these visionaries’ positive outlook?
    • Used to be trying to salvage a sclerotic democratic system
    • Was not capable of imagining more than incremental changes (imagination atrophy)
      • But so is the case of the democratic renovation project
      • US lagging behind many countries in the way it practices democracy
    • Thinking aspirationally about the future can be affirming
    • Being good ancestors
      • Well-being of Future Generations (Wales) Act 2015Counselors of future
      • Wales future generations commissioner
      • Europe has a network of new ministers for figuring out new democracy
      • UN has appointed a special envoy (our common agenda 2021) and youth office
    • How to operationalize these ideas
      • Intergenerational fairness assessment tool
        • Portugal adopted this, the UK is considering it
      • How we give voice to future people
        • Roleplay
          • Future design in japan
      • Strategic foresight capacity
        • Range of tools to identify future scenarios
        • Anticipate future scenarios, stress test assumptions and policy, imagine desired futures and navigate volatility

    David Perry: University of Minnesota

    • Legal historian
    • Example of complex voting systems in Venice in 1229 with doge voting 20 / 20 for the leader, then choosing a new system of voting to prevent future stalemates
    • 30 electors chosen by lot, then 9 selected, then they vote for 40 people, then 12, then vote for 25, then 9, then 45….
      • Final 41 each proposed 1 candidate, all of whom were discussed, then each votes
      • This actually happened, this is real, for hundreds of years!
    • Voting is medieval, it is a normal thing that medieval people did
      • Normalness of voting as a human behavior
      • What it tells us about the Enlightenment
      • How in the West we have constructed this narrative of Athens voting in antiquity
      • That’s not what the history is, and we need to know what it actually is
    • Venice
      • There were hereditary dukes
      • Populists would proclaim a doge by the ninth century
    • Election of the papacy
    • They loved to write bylaws
      • Medieval university worked like this
    • Not aberrations in the fabric of medieval history
    • Medieval voting systems were secret, clever, sought to prevent voting blocs, and were often constructed to keep power in place
    • Peasants frequently sided with the kings against the gentry
    • Increasingly peasants asked for less control by hereditary rulers
    • Commune: former corporate body
    • Even after communal revolt occurred, shifts seemed to require violence

    Bruce Schneier: Fellow and Lecturer, Harvard Kennedy School

    • Positive ways in which AI can help democracy
      • AI as educator
        • Learning through interactive conversation
        • Notably interacting with candidates, or with issues
      • AI as sense-maker
        • Overcoming bottlenecks, summarizing
        • Not many ways for individuals to reach Congress members; AI can read the letters from constituents and highlight agreement, disagreement, and interesting arguments
      • AI as moderator
        • AI is involved in online debates, facilitating conversation, highlighting agreement and disagreement, muting trolls
      • AI as lawmaker
        • Already used in Italy
        • AI creating nuance
        • AIs writing laws that have loopholes
        • In Brazil, a legislator proposed a bill secretly written by GPT 4 that was debated and passed
      • AI as a political strategist
        • Helping local people running for office
        • Helps them navigate bureaucracy, design website, engage with voters, design mail
      • AI as lawyer
        • Right now it is expensive to sue someone, and that is a socialsignal
        • Goes away if doing the work of a lawyer becomes free
      • AI as cheap reasoning operator
        • Generating op-eds, generating tweets…
      • AI as law enforcement
        • Breathalyzers, computerized speedometers, red light cameras
        • Might end up with laws that are not understandable anymore
        • Can’t see source code
      • AIs as propagandists
      • AI as political proxy
        • Our political preferences aren’t on the ground to be discovered, they’re formed through democracy; we probably shouldn’t allow AIs to vote for us
    • Meta questions
      • Who has agency? Is AI augmenting or replacing agency?
      • Efficiency?
        • In our societies we embrace efficiency
        • Inefficiency is sometimes good; it may be desirable for government to run slow and universal
        • How can we master, and then modify, the tech?

    Questions

    • Distinction between complexity and complicatedness? How does it relate to solving democracy?
      • Central to complexity; democracy is unique because it demands that people accept uncertainty over outcome to have certainty over process. A problem about uncertainty is more than an just aggregation of preferences
        • Bruce: problems are complex
          • We’re not going to solve wicked problems with current tools
          • Tools are not suited to world encompassing wicked problems
        • Jordan: prioritizing the propositional over the participatory
          • Speed of trust; bridge building was a participatory exercise that preceded the development of procedures
          • Shifts: from intelligence to wisdom; from risk to uncertainty
          • How we navigate uncertainty, how we grow through the developmental process of transition
          • Phase transition: we built a paradigm—a set of propositions to describe reality to reduce the semantic distance with the way we collaborate
          • As the number of individuals increases, the technology of paradigm is no longer functional. Engaging with disambiguation becomes a heavier burden than operating in a non-paradigmatic model
          • Requires a shift to a participatory model; from legislative first to adjudicative first model of governance
      • AI as generator of courage? having the courage to take risks without knowing what the outcomes are
    • Politics in the UK almost as complex as here. How to articulate between today’s politics, processes, and the imagined visions that are brought?
      • Suzette: what’s happening in Wales is amazing, and there are attempts at identifying governance models that could work in the future
        • Enough people working on this that they’re making peer groups, we’re starting to deal with the uncertainty of other questions
        • The future can be a powerful depolarizing tool; shared aspirations for the future, more common ground than in the present
    • How to talk about these problems when we don’t have the words
      • Science fiction books typically start with definitions
      • Eugene:
        • In science fiction, these fictive techniques have a goal of defamiliarization: trying to construct a narrative that you want someone to engage with openly
        • You can get a different type of engagement from readers by using sci-fi than by using a didactic or direct narrative
      • Victoria:
        • The role of the arts is to do that bridging
        • Picking one thread of current themes
        • Artists are in the world, reflecting on the currents around them, refracting it through their art, one way or another
        • Dystopia as a concept; the Romans didn’t have this concept
        • Using rhetorical techniques and imagination of the future, showing in a lot of ways
      • Rich engagement with arts and education at length for all people
    • Identity; US becoming more racially diverse
      • Suzette: worked for twenty years on immigration policy
        • Minority becoming a majority; we have no idea!
        • Over generations, people change their identity
        • Cautioning against these projections because they are societal constructs, and these can change
      • Inevitability; Jordan
        • Not sure that homo sapiens will remain a relevant term
        • Indy Johar from London
        • Establishing rights of nature at a constitutional level
        • Who has standing, who can defend nature; indigenous groups have a complex way of tackling this problem; in Hawaii a specific lineage has a responsibility to defend water
        • Indy: AR and VR: AI who you can have an interactive relationship with; instead of objective relationship, you
        • One more level up: identity at its most basic element is a locus of choice, relationality
          • Will be conserved across all possible embodiments
    • Futures
      • Track record of people who do economic forecasts has been terrible
      • By creating shared futures, even ones that are not true, you can galvanize shared imaginaries
      • The future is a non-realm; we need to recognize the fictionality of it
      • Bruce:
        • Book: Ethics for a Broken World by Tim Mulgan
          • Ethics of abundance no longer apply
          • Using a speculative future to illustrate ethics of the present
    • The notion of the future in this room is a shared optimistic narrative, but it has also been used as a tool for fear
      • How to respond to attacks on shared sites like libraries
    • Glen: two common narratives of inevitability (Neal Stephenson style: information will lead us into a virtual world) vs. AI is going to take over everything, either making us comfortable or killing us all
      • Thinking a little more actively about these
    • Inaccessibility of these words and conversations
      • Jordan: using AI to increase it, to explain to us
      • David: using AI to explain to us plain language (in the disability world)
      • Victoria: privilege of education and access
        • Like any other kind of privilege, to be aware of it first, and then to use that privilege to disseminate it to others who don’t
        • Art as a way of doing this
        • Putting these in the words of characters, into plain speech
    • Tantum Collins: AI as educator vs. AI as political decision maker
      • A lot of fears about whether AI is good or bad is around who controls it
        • Notion of a public AI seems powerful
    • Rawlsian notion of our grandchildren
      • Notion of identity, what does it mean for the future
    • Suzette: role of speculative fiction in uniting people that doesn’t have to be jargony
      • What is the life we want to live?
      • How to bring the voice of people into these decisions
      • Not only technocratic, has to come from the bottom up
      • In Wales, the passing of this act was a collective decision
      • Counter the fear about the future with aspiration

    Bruce Schneier January 8, 2024 7:00 AM

    Panel #3: Participation

    Lecture

    Jon Evans: science fiction writer

    • Most tech startups are not actually tech startups, they are anthropology experiments
      • It’s about how people are going to react
      • Most anthropology experiments fail, no matter how brilliant their investors
      • The tech industry (for all its flaws) is today’s primary engine of change because it seeds the world with new experiments every year instead of trying to determine in advance which ones will work (ChatGPT, Uber, Airbnb are all anthropological experiments)
        • We have no idea what is going to work!
        • Crucible of petri dishes that it spits out into the world
    • To recreate democracy, we need democratic testbeds
      • No point implementing new aspects or forms of democracy if it turns out that ordinary people don’t care and won’t use them. Trying to “make fetch happen” will waste both resources and people’s willingness to try new things
      • Need to try on a small scale, do experiments
    • Some potential testbeds have existed for a long time
      • Civic groups, fan clubs, churches, intentional communities, professional organizations, gardening societies, any grassroots community who determine (aspects of) their own governance
      • But also special economic zones; how about special democratic zones!
    • Some are very new, and also disreputable
      • DAOs: decentralized autonomous organizations, which already support and experiment with, many innovative kinds of voting…powered and enforced by blockchains
      • CityDAOs, like DAOs but restricted to particular cities, and therefore able to overlap into the real-world community as well
      • “Network state” movements
      • Discords, subreddits
    • Temporary autonomous zones, Physical example: Burning Man
    • New forms of voting
      • Quadratic voting
      • Direct voting, with frequent regular plebiscites
      • Liquid voting
      • Mandatory liquid voting
      • Participatory budgeting: at tax time voters allocate a nontrivial percentage of budget to specific causes
    • New kinds of constituencies
      • Optional online constituencies: people choose to vote either in their geographic district or as a part of a self-sufficient community (“the senator from Reddit”)
    • Forecasting
      • Forecasting works surprisingly well because their errors tend to cancel one another out
      • As such, forecasting the effects of various forms of democratic experimentation would likely be quick

    Archon Fung: Professor of Citizenship and Self-Government, Harvard Kennedy School

    • What kind of democracy for our capitalism?
    • Democratic theorist remarks
      • From a book, Empowering Affected Interests
      • Introduction is from Termination Shock, by Neal Stephenson
        • Single individual who starts geoengineering
      • Solar geoengineering project democracy questions
        • Who should have a say, who does, how do they have a say, how should they have a say?
    • How should democracy regulate capitalism?
      • Foil for these remarks: democracy is representative democracy on a nation-state basis
        • Big democratic state regulates big national capitalism
        • Chain of policymaking: citizens → national representative government → laws, policies, regulations → govern firms, civic organizations, philanthropies, individuals… → allocation of capital, distribution of income and wealth, and so on.
      • Some limits of the national popular sovereignty view
        • Domination of state by capital(ists): political inequality and capture…
        • Limits of state capacity and regulation
        • Changing firm structure—“fissured workplaces” (David Weil)
        • Globalization—of production, capital, labor
      • An alternative: principle of affected interests
        • An individual should be able to influence an organization if and only if that organization makes decisions that regularly or deeply influence that individual’s important interests
        • Instead of a foil, it’s a very different picture: people should have a say in many things that have an influence on their lives (national government, local government, universities)
      • Example with Claudine Gay and Harvard
        • Letter by 700 faculty to retain Claudine Gay
        • They felt they had some right to influence (example of affected interest—the professors are here the longest and they will be the most affected)
        • Different from representative idea!
      • Three applications
        • Workplace and workers (all affected principle)
          • Workers should exercise influence over decisions at work that influence their interests
          • Direct vs. mediated, active vs passive
          • Direct and active: workers bargain with employer, or workers govern the firm
          • Direct passive: workers less hostage to labor market due to social welfare provisions
        • Development NGOs and beneficiaries
          • Beneficiaries should exercise influence over the global philanthropies and development NGOs who affect their important interests (example: Who benefits from Gates Foundation programs?)
        • Current reality: capitalist, nation-state democracy
        • Implications of all-affected principle

    Nils Gilman: SVP of Programs, Berggruen Institute; Deputy editor, Noema

    • Instability of the liberal subject
    • Idea of citizenship that is basically a myth (it should be unitary and every citizen should have exact same rights and responsibilities)
    • Against normative claim
      • Under feudalism, different rights and responsibilities
      • It’s xenophobic: either you’re in or out
      • National sovereignty makes it harder to care for externalities beyond nation
    • Segmentary citizenship
      • What if these rights were divisible
    • Liberal ideal of citizenship aspires to egalitarianism between those included
    • Examples
      • US: nationalistic and has ideas of sovereignty
        • Green card, or H1B visa; undocumented workers, DREAMers Segmentation of rights or responsibilities (de jure segmentation)
      • China has similar mode
        • Hukou: you can work anywhere but only collect social benefits in your home region
      • UAE: different rights given blue collar, white collar
    • You can imagine a more positive version of segmentary citizenship
      • If you’re a citizen of Italy, you can work anywhere in the EU, work for the EU, but nationally only in your country
    • What if you could give segmentary citizenship to neighboring countries
      • Ukrainian citizens effectively have temporary citizenship at the moment
      • Could use this type of citizenship to do political inclusion, be more flexible, could be used for climate migration

    Kathryn Peters: Co-founder, Democracy Works

    • Operations geek, logistician
    • Spent ten years here doing Democracy Works (built out of HKS)
    • Getting rid of legislative elections
      • Extremism, gridlock, unequal representation and influence, interest in running and skill in campaigning are not good predictors of governing ability
    • Alternatives:
      • Sortition (randomly select people to be representatives)
      • Excellent opportunity for academic collaboration as experts and advisors
    • Some are already doing it
      • Citizen’s initiative review in Oregon
      • Redistricting commissions in Colorado and Michigan
      • Advisory bodies to government
      • Advisory bodies to tech companies
    • Why isn’t this more prevalent?
      • 89,000 municipalities, 13,000 school districts
      • Why not start randomizing many of these
    • Can they really take on the hard issues?
      • Past assemblies were mostly invited by authorities to advise
      • We don’t need permission to gather and form citizen’s agendas on hard topics
      • Excellent opportunities for academic collaboration as experts and advisors
      • Why not convene and learn?
    • Whose permission are we waiting for, really?

    Glen Weyl: Economist, Microsoft Research

    • Plurality: intersection of digital and plural
    • Technology vs. democracy
      • Government support in R&D spending is falling
    • Lots of people believe that technology is actually constraining democracy
    • No pure evidence that democracy sustains technology
    • Political ideologies
      • Libertarian story: technology frees everyone, Snow Crash kind of worlds
      • AI kills us all or saves us
      • Plurality: an alternative
    • Plurality in Taiwan
      • Have done this with government
      • Audrey Tang: world’s first transgender minister
    • Writing a book on plurality, the future of collaborative technology, and democracy
      • Technology to facilitate collaboration across social differences
      • Infinite diversity and infinite collaboration (inspired from Star Trek)
    • Three-part meaning
      • Positive: diverse, intersecting groups, not individuals and society
        • Hannah Arendt
      • Normative: diversity is fuel and collaboration across it is engine on which societies thrive
        • Danielle Allen and her work on a connected society
      • Prescriptive: technology can and should foster, empower diversity, collaboration across it
        • Audrey Tang
    • Spectrum of collaboration across diversity
      • Tension between diversity and collaboration: hard to work deeply with people very different from you
      • Point of plurality is to push back this trade-off
        • There are ways to quantify this using information theory
    • Ways to contribute: art, tech companies, academics

    Riley Wong: Research Scientist, Metagov; Independent Researcher

    • Did community organizing; machine learning engineer at Google
    • Privacy enhancing technologies: PETs
      • Zero-knowledge (ZK): prove you know something without revealing the information yourself
      • Multi-party computation (MPC): can run computation on multiple private inputs
      • Fully homomorphic encryption (FHE): everything stays encrypted throughout the entire process
      • And more: differential privacy, federated learning, verifiable credentials, decentralized identity, hardware enclaves
    • Data collectives
      • How do these different infrastructures inform practices around governance, membership, access, and usage, and vice versa?
      • Examples: data co-ops, trusts, coalitions, unions, commons
    • Interfaces for data consent
      • Licensing models—Creative Commons, “copyleft,” and open-source licensing (GPL 3)
      • GDPR cookie consent interface
      • “Consent profiles”
    • Models beyond ownership
      • Land ownership as a colonial import
      • Public infrastructure as examples: public transit, public parks, libraries, archives
      • We don’t need to own public infrastructure
      • How might these models extend to other goods and services such as housing, food, education?
    • Access, usability, and responsibility over ownership
    • Stewardship, consent, and agency over control
    • Community research and co-design
      • Designing by and with communities
        • Tech solutionism, saviorism is quite common in tech

    Questions

    • Meta governance and preconditions
      • There are conditions in international law that structure the kinds of governance that are possible
      • Kathy: many barriers indeed, but invitation cuts through this
        • Inviting the people around you to take part, to vote
      • Riley:
        • What leads to good governance: trust, collaboration…
        • In Taiwan: how does being an island, being relatively homogenous, having a very safe community, having healthcare, and so on, contribute to trust?
      • Jon:
        • Preconditions: money, interest, resentment
    • Dimitri:
      • The more powerful and wealthy I am, the more affected interest I have
        • Archon: that is the principle of corporate governance
          • Doesn’t seem problematic if you do have a stake
        • Glen: What if you think of equality beyond mathematically equal political equality?
          • You don’t want gig workers who work for twenty companies to have more political power
      • Initiatives like in Belgium, which did not get picked up by governments
        • Money, time, motivation, sense of impact and purpose, resistance from the powerful

    Bruce Schneier January 8, 2024 7:01 AM

    Panel #4: Institutions

    Lecture

    Sorcha Brophy: Assistant Professor of Health Policy & Management, Columbia University

    • Drivers of health inequality in the US
    • Imagination in healthcare very structured by current realities of healthcare (e.g., students asked to imagine future healthcare immediately recreated insurance, and then created equity committees
    • FQHC—Federally Qualified Health Center
      • Medicare and Medicaid doesn’t reimburse hospitals 100%
      • FQHCs get enhanced reimbursements per patient that covers the cost of care
        • Enables them to deliver better care
      • Can cover costs of most basic medical services but hard to fund other programs
        • Need to apply for grants—these programs typically don’t get renewed
        • Don’t have room to increase salaries and compete with hospitals
      • Hard to collect metrics they need for reimbursement
      • Limitations they face outside of providing basic services
      • Held back by the imaginative limitations present in our system
        • NY FQHCs describe that funding streams are available for some things but not others (e.g., money is very programmatic, so can apply for specific short-term, grant-limited opportunities and are not renewed, making it hard to have long-term investment
      • Don’t have money to allocate to operational costs
        • no room to increase salaries to keep up with competition, making it impossible to compete with hospitals/clinics, causing short-staffing and burnout
      • Generally limited health record systems
      • Limitations in providing services outside of basic healthcare services
      • Reimbursement challenges mean that individuals have a hard time drawing specialists to work at these clinics
      • Can only provide some services for patients designated by the governments as eligible (e.g., patients with HIV)
      • Makes it very hard to provide comprehensive care for categories of health that there is bipartisan support for—good example of ways to start thinking about healthcare issues
      • Physical infrastructure
      • Legacy systems make it harder to collect quality metrics

    Joshua Fairfield: Professor of Law, Washington and Lee School of Law

    • How to talk about things we haven’t addressed yet, to find the words
    • What’s happening to precedents in courts in the context of AI
    • How to create the kind of language we need
      • Language is humanity’s superpower—lets us massively parallel process new problems and generate new ideas
    • AI systems are eating their own tails, deviating from the common law, and so on
    • Language is humanity’s superpower; allows us to parallel process
    • Humans evolve in historical time unlike ants and other species that evolve in evolutionary time
    • Law is our common language
    • Context matters, even if the words are identical; cannot make the mistake that the combination of community, text, context is not the meaning itself
      • AI doesn’t replace humans in this sense
    • What happens to our ability to massively parallel process our experience (democracy) when we hand it over to generative AI?
      • We construct democracy by un-asking important questions—not tackling them
        • For example, if you don’t answer the question “How much time should children work?” you un-ask it
    • How courts are beginning to use AI
      • Not just a meliorative task
      • The common law is a map of meaning; this case is like that case, because humans have thought there is a link
      • That heatmap of meaning is being replaced; CoCounsel, just launched, isn’t just about “give me a case that says this,” it’s more “give me all the cases that say what I mean”
        • The problem is that these lists become the norm
      • We are moving to an adjudicative system; one in which humans are no longer barometers for what the letter of the law should be

    Misha Glenny: Institute of Human Sciences Vienna

    • The politics of next year, 2024
    • 39 other elections next year excluding the US one!
    • Somewhere deep is a pull for democratic legitimacy
      • 66–75% of Europeans still say democracy is the best system
      • Most people in Vienna and Germany still believe democracy is the best way to do things
    • Simultaneously surging of far right political groups, populist, and so on, and they have become much more efficient than they used to be
      • Candidate for Freedom Party in Austria could become the chancellor—Herbert Kickl
      • They are coordinating an assault
      • They hit four themes: COVID-19, Ukraine (why are we supporting it, make peace with Russia), Energy (through peace with Russia), Immigration
    • Fascism is no longer inconceivable
    • Countries that have experienced with populism have been disillusioned
      • The UK has done this; the Tory party is destroying itself in front of our eyes
        • 60% of the UK population now believes that Brexit was all a big mistake
      • Poland: got rid of the populists quite fast
      • None of the populist countries are pushing for an exit from EU
      • 2004: beginning of the biggest migration to the UK ever; Poles
        • Notably to small towns
      • Bosnia has lost 10 to 20% of its population!
    • What Viktor Orban understood: as soon as you get in, undermine the courts, the media, cultural institutions, before you go back to the ballot box

    Tantum (Teddy) Collins

    • Was previously at DeepMind
    • Stockfish is to AlphaZero as the US Constitution is to what?
      • Stockfish was the highest-performing chess engine pre–machine learning
      • Alpha Zero was created by Google, Deepmind is the update
    • Good old-fashioned AI is much more aligned with what we think about most problem solving and traditional coding
      • breaking down large objective into sub objectives
    • GOFAI vs. ML
      • GOFAI: reductionist, deductive, interpretable, discrete, complicated
      • ML: interdependent, tacit, nuanced, continuous, complex
    • What if we mapped what a cat is by ML—more rigid because a cat is nuanced and has exceptions
    • Tacit knowledge: things we know how to do but we don’t know how to convey in enough granularity to encode effectively
      • Machine learning can
    • Michael Polyani: “We can know more than we can tell”—tacit knowledge
      • Generative AI systems operate in a highly complex space
    • Institutions
      • Rules vs. standards
        • Rule: you cannot drive at more than 55 mph
          • Ease of replicability
        • Standard: you cannot drive recklessly
        • Trade-off between nuance and consistency
          • What if ML for normative representation could expand this frontier?
    • Even if we could do this
      • Bias
      • Input at multiple levels of abstraction and engage in reflective equilibrium
      • How would citizens be informed?
    • Broadly, the limits of interpretability
      • Even though in theory we could understand the tax code, most people don’t
      • The limits of a kind of interpretability; what it would mean to engage with these in a deeper way

    Ulrich Wagrandl: Vienna University of Economics and Business

    • Teaching constitutional law
    • Background in militant democracy
    • Democratic rights can be used against the system
      • You can use freedom of speech to argue against freedom of speech
      • Militant democracy: we will not accept this paradox, we will outlaw it, forbid people to use their rights in an anti-democratic manner
      • In Europe, this is quite common: there are parties that are outlawed based on their ideology—in Germany, Nazi and Communist parties; Fascist in Italy
      • Come from a lived experience of dictatorship
      • Militant in the sense that it is not neutral
    • What kind of democracy we are defending?
      • Liberalism: the rule of rights, constitution
      • We have a democracy limited by its liberal part; symbiosis of two different principles: liberalism (individual rights, autonomy of individual, rule of law, constitutionalism) and democracy (popular sovereignty)
      • Militant democracy defends liberalism; attack on the liberal part is coming from the democratic part
    • The attack that the democratic part is staging on the liberal part is not wholly unjustified
      • Liberal democracy is a limited democracy, constrained by a range of mechanisms, all of which are counter-majoritarian
      • Rule of law that gives final say over governmental action
      • Our institutions are aristocratic or epistocratic—this is a valid charge
        • Don’t embody the democratic ideal; houses of parliament not elected by the population
        • Independent administrative agencies—lots of rule-making power but limited democratic accountability
        • Supreme courts or constitutional courts
        • Proliferation of independent rule-making agencies, not under enough democratic accountability
        • But of course this is for good reasons; power needs to be constrained

    Questions

    • Do we need to change our understanding of interpretability; we’re being asked to do that by very powerful forces; but shouldn’t we change this to “should accept norms of interpretation”?
      • AI closes out the whole tradition of embodied hermeneutics
      • Teddy: we can look at domains where we understand things and see how it has evolved
        • Some time ago people knew how what they were using had been built, how to deconstruct and reconstruct it
        • In some areas we had a fundamental understanding of it, but instead we now delegate our trust to others
        • With AI systems we’d migrate to a new kind of understanding: inductive rather than deductive
      • Joshua:
        • “All men are created equal”: we make it true because we create systems of meaning; we have evolved a normative understanding of this (through law, notably)
        • What AI does, even ML models—it reads our past, not our future
        • If we lose the capacity of saying what is good for us, we lose the ability to create a normative world
    • Limitations of democracy; but what would it mean for the system to be really good? (outcomes?…)
    • Democracy without liberalism? What are we defending?
      • Ulrich:
        • Separate the concepts
        • Democracy collectivist definition: popular sovereignty defines it
          • It is a viable option but not the one liberalism advocates for
      • Misha:
        • What we have under illiberal democracy: autocrats; what we have under liberal democracy: billionaires
    • Issue of immigration, very prevalent whether we want it or not
      • Misha:
        • Elections next year include Taiwan, India, Indonesia…
        • Migration plays a large role, especially in India
        • People in Europe recruiting people in West Africa to work in nursing homes and so on
        • Although everyone knew migration was a key issue in Brexit, nobody focused on it
        • Solution to this requires a huge investment into regulatory systems that don’t currently exist, in the US, UK…
          • Need a system that reduces the humiliation of migrants
          • Requires courage from political elite
    • Helene:
      • Replacing constitution with a piece of code
      • Nobody can be excused by ignorance of the law, this is a basis of law; we couldn’t do this if we don’t even understand the law
      • Teddy:
        • Define values based on people’s inputs: in a scale-free way, enable people to input their judgment
        • Along one dimension could be very democratic
        • The function that takes these inputs and maps them onto a desired outcome; not 100% clear to him that this is less legible for most people
    • How our existing frameworks limit our capacity for imagination
      • Joshua:
        • There is a fundamental difference in making meaning together
        • By design, AI will eliminate the outcome
      • Teddy:
        • It is true that most AI systems are trained on historical data, but that doesn’t mean that they can’t devise creative solutions to problems
        • Could AI generate law?
          • Usually we think of it in the context of humans writing words
          • Maybe there’s a possibility for the representational medium to be different than text
      • Joshua:
        • AI can’t create what’s new because what’s new is defined by humans—we give it meaning; otherwise, it’s just a stochastic series of operations
        • “New” only matters because we decide it’s new
        • Hallucinations is when it loses its grounding in context
      • Institutional press
        • Does it have a role?
        • Misha:
          • Two fundamental assaults on media
            • From money: Fox in the US, much of the British mainstream media
              • Suggests that there’s still value in it politically
            • From public society, disinformation
              • There are very cheap ways to influence political processes with Facebook, TikTok
              • Putin has put so much effort into disinformation
          • Distinguishing between original and reported news
          • Focusing on education from the very beginning
        • Tasha:
          • Putin wants to have some democratic appearance because there is performative value
      • Aviv:
        • Case law grounding for AI systems
          • Social futures lab
          • If you even get a set of people to do content moderation based on a set of rules, they do less well relative to if you give them a set of precedents and let them pick which one is relevant
          • Actually gets closer to what the humans could have done; using precedent as a mechanism

    Bruce Schneier January 8, 2024 7:02 AM

    Panel #5: AI

    Lecture

    Judith Donath: Faculty Associate, Harvard University Berkman Klein Center

    • Biology:
      • Signaling theory in evolutionary biology:
        • Recent developments in evolutionary biology
        • Unexplained phenomena, like the peacock’s tail, can be guarantees of reliability
        • Examples from economics: displaying abundance of a specific resource
      • Foundations of communication from a design perspective:
        • Discusses the design perspective of communication
      • Governance and trust:
        • Examines the role of governance in building trust
        • Considers the concept of surveillance in human societies
          • Gossip as a form of surveillance (Dunbar)
          • Linguistic relation between gossip and religious notions
          • Divine punishment and surveillance as factors enabling the growth of human societies
          • Humans’ belief in constant surveillance contributes to societal cohesion (religion provides the sense that you’re being watched all the time)
    • AI:
      • Rational and irrational fear of AI:
        • Explores the irrational fear of AI, especially among its developers
          • Elon Musk’s cynical view
          • Human desire to control or develop AI
          • The potential for weaponization of AI technologies
        • Reflects on the motivation behind entering the field of machine learning
        • Draws parallels between historical idol worship and beliefs about building AI systems
      • AI as a twenty-firstst century version of an omniscient god:
        • Examines the key question of AI systems and our belief in them
        • Surveillance technologies in governance, comparing them to religious surveillance
        • Examples: Uber as a surveillance mechanism
        • Considers potential future governance systems and their means of operation
    • Connecting themes:
      • Role of surveillance:
        • Surveillance as a fundamental aspect of both biological and AI systems
        • Drawing parallels between religious surveillance, societal surveillance (gossip), and technological surveillance
    • Conclusion:
      • Suggests a connection between AI development and the concept of a god, with surveillance as a central theme
      • Raises questions about the legitimacy and means of future governance systems, drawing connections between AI and divine attributes

    Iason Gabriel: Staff Research Scientist, Google DeepMind
    Topic: AI, Power, and Legitimacy

    • Part 1: Understanding the power problem:
      • Background:
        • Iason Gabriel, a philosopher associated with DeepMind
        • Talk title: “A New Leviathan”
      • Business as usual and the new leviathan:
        • Companies and AI shaping information and preferences
        • The power problem: companies shaping information and preferences
        • AI embodies the principles of its creators, leading to the “value-alignment” problem
        • AI systems becoming increasingly agentic
      • Legitimacy question:
        • Is the current situation legitimate?
      • Value problem:
        • Large generative models
        • AI alignment rules and the challenge of legitimacy
        • Where did the rules come from?
        • The source of rules: human preferences and compute workers (often not representative)
    • Part 2: Proposed solutions:
      • Disagreement on values:
        • The need for a process or principles not dependent on moral suppositions
      • Option 1: Ideal spectator or Rawlsian veil of ignorance:
        • Global exercise
        • Regulations, stringent safety standards, and protection of human autonomy
        • Equal access to opportunities created by the technology
        • Material dividends to benefit all evenly, with more benefits for those least well off
      • Option 2: Democratize AI:
        • Involving citizen assemblies and annotators as representatives of communities
    • Part 3: Reflections on solutions:
      • Constitutional process:
        • Discussion on the power of representatives in a constitutional process
      • Legitimacy requirements:
        • Justification and consent as essential elements
      • Democratic AI consideration:
        • Evaluation of whether to pursue democratic AI as suggested
    • Connecting themes:
      • Power problem:
        • Addressing how companies and AI shape information and preferences
      • Legitimacy:
        • Exploring the legitimacy of AI development, focusing on value problems and outsourcing
    • Conclusion:
      • Suggests potential solutions, emphasizing the need for a thoughtful, global process or principles to ensure the legitimacy of AI development
      • Raises questions about democratizing AI and the power dynamics involved

    Paul Gölz: Simons Laufer Mathematical Sciences Institute, Cornell University
    Topic: Generative Social Choice

    • Introduction:
      • Paul Gölz focuses on social choice theory, a mathematical theory of voting
      • Choosing between few alternatives (for example, Brexit is a binary decision)
      • The objective is scalable, open-ended social choice
    • Challenges:
      • Identifying unforeseen alternatives
      • Generation of alternatives
      • Handling unknown preferences
      • Discrimination between ideas
    • Framework for social choice processes:
      • General framework to guide the design of social choice processes
      • Two components:
        • Guarantees with perfect queries
        • Empirical validation of queries
          • Enables exploration of empirics for real-world application
    • Application domain: representative statement selection:
      • Objective:
        • Develop a slate of k statements that satisfies representation axioms from committee elections literature
      • Process:
        • Input: free-form opinion expressions
        • Output: slate of k statements representing heterogeneity, not just a single decision
        • Slate should satisfy representation axioms from committee elections literature
      • Pilot on chatbot personalization:
        • Applied the method to the domain of chatbot personalization
        • Procedure:
          • Democratic inputs to OpenAI’s AI project
          • Generation and validation of statements
          • Took a representative sample of 100 Americans stratified by categories and surveyed them. Then ran algorithms and generated 5 summary statements and then got 100 new participants and asked them how well the statements summarized how they felt
          • Evaluation of how well participants feel represented
            • 75% felt perfectly represented
            • 18% felt mostly represented
    • Generative social choice:
      • Choosing between few alternatives, with emphasis on open-ended deliberation
    • Challenges revisited:
      • Identifying unforeseen alternatives through generation
      • Handling unknown preferences and discrimination
    • Conclusion:
      • Paul Gölz’s method, generative social choice, demonstrates a framework for handling unforeseen alternatives and unknown preferences in social choice processes
      • Empirical validation, as shown in the chatbot personalization pilot, contributes to the method’s practical application

    Helene Landemore: Department of Political Science, Yale University
    Topic: AI, Democracy, and Corporate Governance

    • Introduction:
      • Helene Landemore from Yale University, a political theorist
      • Focuses on the potential for AI to enhance deliberative democracy and rethinking corporate governance
    • Part 1: AI’s potential in democracy:
      • Scaling democracy with AI:
        • Explores the idea of using AI to scale democracy, particularly in citizen assemblies
        • Proposes AI augmentation for tasks traditionally handled by humans
      • Six ideas to augment democracy:
        • Selection and distribution of participants:
          • Use of algorithms (e.g., Lexi) to select participants in citizen assemblies
        • Facilitation: facilitators are important to early parts of deliberative processes
          • Challenges with human facilitators (expensive, bias, uneven, micromanagers), suggesting AI alternatives
        • Translation:
          • Introduces a meta-system called Seamless for multilingual deliberation
          • Acknowledges the need for human translators due to their unique abilities
        • Fact-checking:
          • Controversial idea of AI generating summaries, especially on scientific topics
        • Clustering and organizing arguments:
          • AI’s role in organizing and clustering arguments and opinions
        • Deliberation output summary:
          • Highlights the fatigue of human deliberators and the potential of AI to improve summarization
          • Meta did study and found that LLMs are 50% better than humans at summarizing exchanges
    • Part 2: Rethinking corporate governance:
      • Democracy at the corporate level:
        • Proposes normative considerations for introducing more democracy at the corporate level
        • Challenges the legitimacy of current corporate structures
      • alternative models:
        • Technocratic oversight board (e.g., Meta):
          • Critiques its technocratic nature and dubious independence
        • Deliberative representation of stakeholders:
          • Envisions a citizen assembly representative of all humanity influencing corporate governance
        • Addressing objections:
          • Challenges objections related to profit orientation, investor ownership, and the role of workers and consumers
    • Conclusion:
      • Helene Landemore suggests that AI can transform democratic processes, particularly in citizen assemblies, and that rethinking corporate governance through democratic principles is essential

    Aviv Ovadya: Berkman Klein Center, Harvard University
    Topic: Reimagining Democracy for AI

    • Introduction:
      • Aviv Ovadya from the Berkman Klein Center, Harvard University
      • Aims to reimagine democracy for AI in a world where democracy can keep pace with AI
    • Assumptions:
      • Everyone will be significantly impacted by AI
      • Everyone should have a voice in charting our AI future
      • Lack of voice leads to resistance and compromises safety
    • Current trajectory: unilateral AI:
      • Two flavors:
        • Autocracy: unilateral control by powerful corporations or authoritarian countries
        • Chaos: unilateral distribution of AI systems without guardrails
    • Alternative: democratic deployment of AI:
      • Challenges with existing national governments
      • Introduction of AI platform democracy
    • Global democratic processes:
      • Deliberative democracy and AI-augmented tools
      • Examples: European citizen’s panels
    • Complete AI/platform democracy:
      • Representative, informed, and, where appropriate, compensated participants
      • Making binding collective decisions about policies or values for AI systems across national boundaries
    • A recipe for AI/platform democracy:
      • Representative deliberations:
        • Enables transnational feasibility; increases representativeness, equity, and expertise
      • AI augmentation:
        • Lowers the cost of transnational processes and enables new forms of interaction
      • Democracy as a service:
        • Neutral third-party organizations running democratic or plural processes
      • AI democracy/platform democracy:
        • Utilizing the above technologies
    • Three layers of AI democracy:
      • Transnational or global AI governance
      • AI organization/corporation governance
      • Governance/alignment of corporations
    • Initial steps:
      • Mention of OpenAI, Meta, Anthropic, and DeepMind
      • Acknowledges imperfections but sees them heading in the right direction
    • Research and practice agenda:
      • Questions the representativeness, meaningfulness, and robustness of processes
      • Highlights risks of “participation washing”
      • Emphasizes the need to compare processes to realistic alternatives, not just democratic ideals
    • Acceleration of democratic innovation:
      • Calls for innovators, adopters, funders, evaluators, and executors
      • Urges to challenge the myth that unilateral AI is the only option
    • Conclusion and summary:
      • Reimagining democracy for AI is crucial as everyone will be impacted
      • Urgent need for a transnational democratic infrastructure for AI to avoid unilateral control over our future
      • There may be very little time; we will need a well-resourced field of applied democracy

    Questions

    • Introduction:
      • Collection of questions and discussions on various aspects of AI and governance
    • Democracy vs. governance:
      • Exploration of the difference between democracy and governance
      • Consideration of the need for people to be asked in a transnational environment
      • Emphasis on the enforcement mechanism, accountability, and challenge of global democratic governance infrastructure
    • Facilitators and AI in democracy:
      • Debates on using AI as facilitators and its role in various aspects of democracy
      • The trade-off between human biases and AI capabilities
      • Discussion on the activity of participating in democracy and the potential for AI to enhance the autonomy of deliberators
    • Complexity and global democracy:
      • Tension between flat and complex understandings of democracy
      • Debate on a global polity and its potential harm in imposing a uniform system on diverse communities
      • The urgency of addressing the problem with a simplified view of democracy
    • Governance of technology and socio-technical context:
      • Exploration of historically embedded beliefs influencing technological developments
      • Acknowledgment that technology cannot be divorced from its socio-technical setting
      • Concerns about the fear among those developing technologies
    • Meaningful consent and AI relationships:
      • Discussion on meaningful consent in the context of increasing AI involvement
      • Consideration of ongoing relationships with powerful AI tools and the need to redefine consent in such relationships
      • Reflection on the changing dynamics of consent as AI becomes more powerful
    • Information overload and governance:
      • Awareness of the shift from informational scarcity to information overflow in the age of AI
      • Questions about mechanisms to accurately capture preferences in a world with more perfect knowledge
    • AI as facilitator and sensemaker:
      • The role of AI in facilitating and sensemaking in conversations
      • Consideration of the trade-offs and the need to maintain a balance between AI removing tedious tasks and preserving creative tasks
    • From tool to tech solutionism:
      • Questions about when AI transitions from being a tool to tech solutionism
      • Concerns about assumptions regarding discovering a set of preferences and the impact of the process of uncovering preferences
      • Emphasis on the human element in making preferences come to life
    • Real problems of democracy:
      • Call for a focus on solving acute problems of democracy rather than adopting new technologies without addressing existing issues
    • Digital state infrastructures:
      • Discussion on digital state infrastructures to address biases in AI systems
      • Uncertainty about whether the issue can be solved solely by having a good dataset
    • Conclusion:
      • The need for multilevel forms of governance
      • Acknowledgment of the fear among those developing AI technologies
      • Emphasis on the importance of addressing real problems in democracy rather than being overly focused on AI-related issues

    Bruce Schneier January 8, 2024 7:03 AM

    Panel #6: Trust

    Lecture

    Niclas Boehmer: Fellow, Harvard University

    Topic: Algorithms and Democracy

    • Participatory budgeting:
      • Aim: make people feel they have a say
      • Comparison: unfair simple greedy rule vs. not manually executable, unpredictable proportional rule
    • The power of algorithmic decision-making systems (ADMs):
      • Throughput: objective, more voices heard
      • Input: scalability—accurately captured
      • Decision-making: more accurate
    • Transparency in ADMs:
      • Challenge: difficult due to inherent complexity, black box nature, potential counterintuitive actions, and lack of human trust
      • Importance: undermines legitimacy, loss of trust leads to disengagement, risk of complete failure, system designers hold significant power, unclear interaction with the system, black box leads to a lack of accountability
    • How to achieve transparency:
      • The how (transparency in process):
        • Make decision-making process transparent by publishing code and data
        • Issues: average citizens struggle to derive meaning from code; lacks independent and intuitive verification (even for experts)
      • The why (transparency in reasoning):
        • Present reasoning behind decisions
        • Human decision-making: simple procedures for easy understanding
        • Algorithmic decision-making: decisions with explanations (verification codes) that are appealing, easy to understand, using ADMs as verifiable “oracles,” employing axiomatic methods to satisfy desirable properties
    • Towards a solution: a multidisciplinary effort:
      • Question: How much transparency are we willing to sacrifice?
      • Goal: identify and work towards trade-offs acceptable to society
      • Technologists:
        • Enhance explainability of ADMs (XAI)
        • Put explainability at the start, not the end
      • Social scientists:
        • Identify acceptable trade-offs for society in various contexts
        • Explore how transparency links to the perceived legitimacy of ADMs

    Summary: Niclas Boehmer explores the intersection of algorithms, participatory democracy, and transparency in ADMs. The challenges include the inherent complexity of ADMs and the difficulty in achieving transparency. The multidisciplinary approach involves technologists enhancing explainability and social scientists identifying acceptable trade-offs for societal transparency and legitimacy. The goal is to strike a balance that aligns with societal values.

    Wes Chow: MIT Center for Constructive Communication

    • FORA:
      • Description: FORA is a MIT-developed tech platform focusing providing tools for constructive communication
      • Operation: it brings small groups together, converting audio conversations into text for analysis
      • Deployment: with a five- to six-year field history, FORA handles six to eight participants in 1.5-hour conversations, serving millions
      • Development: engaged 12+ engineers, leadership, partnership, and support teams with 13,000 commits and 100k+ lines of code, 10k+ participants
    • Project Odessa:
      • Description: Project Odessa is a decentralized social systems app, a more limited version of FORA
      • Features: involves persona management, allowing different presentations of self in various contexts (similar to a finsta, or fake Instagram account)
      • Team: smaller scope with four engineers, two designers, and graduate researchers. 300 commits, <10k lines of code
      • Persona management: Different profiles have different presentations of self (like a finsta)
      • Moderation: implements community-based moderation rules
    • Context and context collapse:
      • Real world vs. social media: discusses context and context collapse, comparing real-world intimacy gradients to the flattening of context in social media
      • Addressing context in FORA: FOR A addresses context and context collapse by providing different visibility settings

    Summary: These projects emphasize the importance of technology in facilitating constructive conversations, with FORA’s extensive deployment history and Project Odessa’s focus on decentralized social systems. The context and context collapse discussions highlight the challenges in maintaining meaningful interactions, both in the real world and online.

    Sam Gill: Doris Duke Foundation

    • Background:
      • Helped build the Knight Research Network
      • Currently heads the Doris Duke Foundation
    • Enlightenment overview:
      • Pre-Enlightenment: revealed knowledge or coercion legitimizes institutional power
      • Post-Enlightenment: objectively derived knowledge legitimizes institutional power
      • Shift in moral status determination
    • Liberal institutions (5th grade version):
      • Goal: conjoin objective power with knowledge
      • Components: objectively derived information, technical expertise, repeatable method = authoritative knowledge
      • Elements: auditability + external constituents = accountability. Publicly assigned normative end = representation
      • Liberal institution = (knowledge + accountability) * representation
    • Issues with social technology application:
      • Rational normativity: surveillance, coercion, reduction
      • Liberal normativity: violence, exploitation, marginalization, powerlessness, cultural imperialism
    • Contingent and obsolete features in liberal constructs:
      • Liberal subjectivity: universalism coextensive with geography
      • Liberal normativity: prosperity coextensive with growth in consumption
    • Challenges for liberal institutions:
      • Temporally broad problems
      • Spatially broad problems (including struggles in addressing)
      • Domination problems
    • Trust issues:
      • Types: transactional trust (I believe the terms of this exchange will be fulfilled), social trust (I feel safe), institutional trust (I believe this institution serves my interests), elite trust (I believe treating non-elites as ends in themselves serves my interests)
      • Expressions and absence of each type
    • Rewiring liberal institutions—four examples:
      • Medicine (enterprise):
        • Liberal flaw: racism
        • Rational flaw: naturalistic fallacy
        • Intervention points: addressing race in clinical equations, race-conscious medicine
      • Child protection (administrative system):
        • Liberal flaw: racism
        • Rational flaw: high modernism
        • Intervention points: investing in prevention-based services, “family well-being” system
      • Conservation (regime):
        • Liberal flaw: racism
        • Rational flaw: logocentrism/extraction/property
        • Intervention points: supporting indigenous-led conservation, integrative environmental paradigm
      • Performing arts (enterprise):
        • Liberal flaw: underinvestment in a public good
        • Rational flaw: category confusion
        • Intervention points: demonstrating non-instrumental benefits, public/private expenditure commensurate with public value

    Woodrow Hartzog: Professor of Law, Boston University

    • Introduction:
      • Emphasizes two key points: (1) powerful actors in our society (big tech) routinely and constitutionally betray our trust at the expense of democracy, and (2) to protect democracy lawmakers must move on individual models of trust to make powerful actors loyal to those who trust them
    • Powerful actors and trust betrayal:
      • Platforms have large incentives for unethical behavior
      • Adversarial outcomes include loss of privacy, degradation of mental well-being, and misinformation
      • Results in societal distrust and isolation
    • Ineffective approaches:
      • Individualistic approaches are inadequate
      • Consent is problematic:
        • Illusory agency due to limited meaningful choices
        • Excessive and overwhelming consents
        • Myopic focus on systemic issues rather than individual decisions
    • Trust as a marketing tool:
      • Trap: attempts to manipulate individuals into exposing themselves
      • Mandatory: imposed as a prerequisite for various activities, eliminating genuine choice
    • Proposed solution:
      • Recognition of millions of created relationships with companies and individuals
      • Apply legal principles of duties of loyalty and acting in the best interest
      • Implement subsidiary rules to minimize flexibility
    • Advantages of proposed solution:
      • Highlights social values over individual preferences
      • Shifts focus from harm to the degradation of the relationship itself
      • Loyalty becomes an anti-betrayal ethos
    • Protecting democracy:
      • Advocates moving beyond individualistic remedies
      • Shifts focus from individual models of trust to ensuring loyalty from powerful actors to those who trust them
    • Addressing platform incentives:
      • Platforms have inherent incentives for self-dealing with detrimental outcomes
      • Issues include attention theft, degradation of mental well-being, and labor theft
    • Reconceptualizing consent:
      • Consent is inherently flawed:
        • Lack of genuine agency
        • Overwhelming nature of consents
        • Myopic output of billions of self-motivated decisions
    • Legal opportunity:
      • Millions of relationships between companies and people
      • Legal approach based on lopsided relationships: duty of loyalty, care, and confidentiality
    • Alternative legal regime:
      • Mandate loyalty and acting in the best interests
      • Focus on collective interest
      • Implement specific rules with limited flexibility
    • Advantages of new legal regime:
      • Emphasizes social values over individual preferences
      • Shifts focus from harm to the degradation of the relationship itself
      • Addresses the vulnerability inherent in power dynamics within these relationships

    Francesca Tripodi: UNC Chapel Hill

    • Author: The Propagandists’ Playbook
    • Information trust:
      • Dominantly sourced from Google
      • Discussion on the impact of the #StopTheSteal misinformation leading to the Capitol attack
    • Perception of truth:
      • Example: agreement on the color of the sky as a shared truth
      • Acknowledgment that truths can vary historically
    • Search engine influence:
      • Preconceived notions shape search results
      • Introduction of “deep stories” (told in Church, communities, campfire) in sociology, emotionally convincing but not necessarily grounded in reality
    • Historical perspective:
      • Mention of the 1870s Reconstruction Act
      • Reference to W.E.B. DuBois’ insights on misinformation during the 1935 Black Reconstruction
    • Keyword curation and strategic signaling:
      • Propagandists utilize deep stories through keyword curation and strategic signaling
    • IKEA effect of misinformation:
      • Trust in misinformation increases when individuals participate in creating it
      • Libraries and librarians, traditionally trusted, are challenged by the trend of encouraging independent research
      • Introduction of “participatory disinformation” with the slogan “do your own research.”
    • Data void and political manipulation:
      • Concept of a data void exploited for political purposes
      • Critique of the current “fix” mentality towards disinformation, emphasizing its intentional use as a political tool
    • Deeper understanding needed:
      • Challenge to view disinformation as a feature for political gain rather than a coding error
      • Call to think more deeply about the underlying narratives and “deep stories.”

    Questions

    • Aditi’s inquiry:
      • Exploration of explanations in elections, considering complexity
      • Current limitations of journalist understanding and potential scalability of alternative systems
    • Tantum’s concerns:
      • Addressing disinformation and the balance between self-trust and deferring to expertise
      • Francesca’s insights as an ethnographer, highlighting the role of faith in conservatism and media literacy
    • Archon’s differentiation:
      • Differentiating between trust and trustworthiness
      • Proposing a need for fewer propagandists and the potential positive use of the IKEA effect
    • Dimitri’s exploration:
      • Examining conspiracy theories on the left and the potential existence of deep stories
      • Woodrow’s perspective on creating spaces to contest disloyal actions by platforms
    • Paul’s IKEA effect question:
      • Inquiry into how media organizations can better communicate using the IKEA effect
    • Bruce’s transparency and trust:
      • Reflection on transparency, cryptography, and the challenge of examining practices versus relying on trust
    • Wes’s technological role:
      • The importance of technology in bridging community divides based on shared values
    • Eli’s values and democracy:
      • Advocacy for considering shared values and the emotional aspect of democracy
    • Automated aid for narratives:
      • Exploration of using automated aid to discover and study narratives not surfaced by human conversations
    • Deb’s insights:
      • Reflection on trust post-Trump’s election and the role of source alignment with individual values
      • Querying the Washington Post editor on transparency regarding story selection

    Bruce Schneier January 8, 2024 7:04 AM

    Panel #7 Deliberation

    Lecture

    Claudia Chwalisz: DemocracyNext

    • Spaces for collaboration:
      • Focus on creating spaces maximizing collaboration beyond traditional democracy paradigms
    • New paradigm of democracy:
      • Suggestion to move beyond the traditional concept of democracy
      • Emphasis on new institutions, processes, rituals, and spaces centered around deliberation and rooted in relational dynamics
    • Deliberation and more than human:
      • Deliberation involves weighing evidence and reaching shared decisions
      • Recognition of more than human, reconnecting with the natural world and understanding our relationship with it
      • The disconnect we have between us and the natural worlds makes us feel more isolated in our daily lives. Autocrats are taking advantage of this trend
    • Citizen assemblies:
      • As a form of deliberate space and anchored in a view of the world of who has agency and rights is a key part of the change we need and a new paradigm of democracy with new institutions
      • Example of 184 people in France meeting for end-of-life deliberations
        • Met with experts for 24 days to talk about end-of-life care
        • Achieved 92% consensus on 67 recommendations
      • DemocracyNext’s effort to systematize deliberation knowledge with MIT’s Citizen’s Assembly Lab
    • Career questions:
      • Twelve guiding questions for her career, including shifting decision-making power to citizen assemblies, ensuring accountability, and exploring the role of technology
    • Humanity and natural world:
      • Reflecting on the impact of disconnection from the natural world on human relationships
      • Autocrats exploiting the trend of isolation resulting from this disconnect
    • OECD study and permanent assemblies:
      • OECD study on 800 examples of citizen assemblies globally
      • Exploring permanent and anchored assemblies
      • Collaborative efforts with museums, MIT, and constructive communication centers
    • Key questions for assemblies:
      • Addressing questions posed by Richard Fineman, including the shift of power to assemblies, accountability, compulsory participation, and the integration of the more than human concept, decisions on who decides and the enabling process, how to reimagine government systems after conflict and how, how technology can be leveraged to enhance deliberative spaces like citizen assemblies
    • Paradigm shift moments:
      • Reflection on paradigm-shifting moments like the Lehman Brothers crash and the current uncertain era
      • Emphasis on recognizing uncertainty as a call to action

    Dimitri Courant: Harvard Kennedy School Ash Center

    • Typologies for citizen assemblies:
      • Time as a factor of differentiation:
        • Exceptional: deals with rare issues in a crisis or after a major change
        • Occasional: relevance at the arbitrary discretion of decision-makers; the Decision-Making Process (DMP) codified as an institution
        • Event-based: predefined events (election, referendum) systematically associated with a deliberative institution
        • Regular: own regular rhythm rather than connecting a DMP to another event
        • Permanent: a permanent institution with clear and fixed prerogatives; recurrent calls for a chamber
    • Power to the mini public:
      • Three key actors:
        • People or citizenry at large (referendum)
        • Rules or elected officials (elections)
        • DMP or randomly selected officials (sortition)
      • Focus on two powers:
        • Agenda-setting
        • Ratification (approval or rejection of DMP’s propositions)
      • Three pure models:
        • Tamed consultation (within electocracy):
          • New metamorphosis of representative government, deepening audience democracy
          • Minimalist (purely consultative citizen assemblies) to maximalist forms (permanent DMP as second, subordinate chamber of parliament)
          • Enlightened counterfactual DMP (e.g., gilets jaunes protests, Brexit)
        • Radical democracy:
          • Extension of a system of direct democracy
          • Minimal (CIR in USA and Switzerland) to maximal forms (DMPs work on policy)
          • Vision: elected officials have failed the people
        • Representative klerocracy:
          • Delegative system, representatives selected by lot rather than election
          • Shared distrust of people, activists, and elites
          • Denies “consent of the governed”
      • Complexifying institutionalization:
        • Different types of hybrid models
        • Pure models: tamed consultation, radical democracy
        • Hybrid: mixed descending model, mixed ascending model, permanent DMP with veto, permanent DMP with agenda setting power
    • Typologies in regard to citizen assemblies (additional points from set 2):
      • Conflicts between institutional models:
        • Institutional design is crucial (people, elected officials, randomly selected citizens)
      • Focus on two powers:
        • Agenda setting
        • Ratification (approved or rejected by DMP’s propositions)
      • Tamed consultation (within electocracy):
        • New metamorphosis of representative government, deepening audience democracy
        • Minimal to maximal forms
        • Enlightened counterfactual DMP (e.g., gilets jaunes, Brexit, Abstention)
      • Radical democracy:
        • Extension of a system of direct democracy
        • Minimal to maximal forms
        • Vision: elected officials have failed the people
      • Representative klerocracy:
        • Delegate system, representatives selected by lot rather than election
        • Shared distrust of people, activists, and elites
        • Denies “consent of the governed.”
      • Complexifying institutional models:
        • Pure models and hybrid models (mixed descending, mixed ascending, permanent DMP with veto, permanent DMP with agenda-setting power)

    Dimitra Dimitrakopoulou: MIT Center for Constructive Communication (C3). Tech-enhanced Citizens’ Assemblies

    • Collaboration with DemocracyNext:
      • Seeds of collaboration started a year ago
    • Tech-enhanced citizen’s assemblies:
      • Leads translational research work at MIT C3
      • Two-year exploration plan:
        • Year 1: use existing tools and early prototypes
        • Year 2: extend features and capabilities of tech as a key infrastructure
    • Building blocks of a citizen’s assembly:
      • Community engagement and listening
      • Learning
      • Deliberation and consensus-building with AI sensemaking
        • Bring people together in small circles, record conversations, share experiences instead of opinions
        • Learning portion is huge in citizen’s assemblies and tech can help
        • AI-powered sensemaking to help with deliberation and consensus building
          • Can then feed into the process of drafting recommendations
      • Public archive of deliberation grounded in voice
    • Community dialogues model:
      • Participant-led small groups (4–6 participants)
      • Prompts designed for open-ended sharing of experiences
      • Recorded conversations for maximum impact
    • AI-powered sensemaking: Cortico
      • Identify key moments in conversations
      • Flowers:
        • Measure conversations
        • Evaluate how conversations went
        • Assess the role of the facilitator
    • Why tech?
      • Consider wider diversity of citizen perspectives
      • Ease the burden of deliberative process (subdivide-synthesize)
      • Ensure a more transparent and accountable process
      • Greater trust and scale
    • Tech-enhanced setup, deliberation, deliverables:
      • Enable and amplify community engagement
      • AI-powered sensemaking and visualization
      • Support deliberation in both small group conversations and plenaries
    • Tech-enhanced citizens’ assembly popup lab:
      • Year 1: use existing tools
    • Miscellaneous
      • Just announced partnership with DemocracyNext
      • DemocracyNext citizen’s assembly guide
      • Tech can enhance community engagement and listening
      • AI-powered sensemaking for deliberation and consensus building
      • Public archive of deliberation grounded in voice
      • Community dialogues model
      • Cortico: video-conferenced conversations with AI-powered sensemaking

    Ruthanna Emrys: Author
    Topic: Building the Garden: Designing Networks to Enable Humanity

    • Popular networking media optimization:
      • Current platforms optimize engagement over:
        • Cognitive comfort
        • Clear collective and individual thinking
        • Democracy
        • Planetary survival
      • Corporate goals prioritize over human cognitive well-being
    • Changing technological priorities:
      • Facilitate focused processing
      • Make scale, priority, and agency salient
      • Build on sensory perception without hindering it
      • Balance action and rest
      • Customize for neurotypes, social styles, and personal preferences
    • Sources of potential:
      • Citizen science and crowdsourced governance
      • Augmented reality
      • Sensory augmentation and substitution
      • Rights of nature
    • One possible direction:
      • Build on crowdsourcing platforms with desirably biased algorithms for long-term values
      • Augmented reality with deeply explorable local environments
      • Sensory augmentation with customizable, multi-sensory interface options
      • Social technology (like democracy) with new ways of organizing life, labor, and joy for a more richly livable future
    • Science fiction as a deliberative laboratory:
      • Explore speculative scenarios:
        • What if fact-checkers ruled the world, defining election rules and frequency?
        • What if all group affiliations were voluntary?
        • What if you could change citizenship at will in five minutes?
        • What if we tried anarchism as a serious system?
        • The biggest question for deliberation
    • Alternative approaches to deliberative speculation:
      • Question assumptions about future development
      • Question assumptions about how knowledge is created
      • Use familiar places to localize implications
      • Consciously imagine our descendants (genetic, social, intellectual, philosophical) living in our imagined futures

    Bailey Flanigan: Carnegie Mellon University

    • Background:
      • Completed PhD at CMU, advised by Ariel
      • Blending computer science, political science, and practice
      • Emphasizes the interdisciplinary nature of the workshop
    • Initial question:
      • Started with increasing inclusion in a computer science department
      • Recognized different perspectives in CS departments: some never included, others always excluded
    • CS JEDI (Justice, Equity, Diversity, and Inclusion):
      • Introduction to justice, equity, diversity, and inclusion in computer science
      • Previous focus on studying citizen’s assemblies algorithmically: how to pick a representative sample
      • Investigating how people internalize the interests of others in deliberation (public-spiritedness, empathy)
    • Measures of internalization:
      • How do people learn about the effects on others?
      • To what extent and in what sense do political preferences account for this?
      • How do people think about risks to others?
      • Whose interests are considered in the outcomes?
    • Factors influencing measures:
      • Underlying similarity/homophily
      • Preexisting biases
      • Existing propaganda/polarization/deep stories
      • Design of information/deliberation
      • Composition of the group
      • Integration of AI
      • Cultural context
      • Ongoing experiments in:
        • Experimental design (pre- and post-deliberation)
        • Classroom deliberation in American politics
        • Deliberative town hall in Chile
        • Citizen’s assembly on energy pricing
        • Social choice theory application

    Hahrie Han: Director, SNF Agora Institute; Professor of Political Science, Johns Hopkins University

    • Mapping the modern agora:
      • Emphasizes the importance of understanding democratic tendencies before, rather than after, an event
      • Civinomics: a classification-based approach to understanding
      • Civil society institutions can be both pro- and anti-democratic
    • Undivided: the quest for racial solidarity in an American church:
      • Focus on megachurches as growing civic entities
      • Example: Crossroads Church—35K weekly attendees, 500K online
      • Average megachurch has 6k regular participants and budget of $5.3 million
      • Leadership and organization/structure matter in these churches
    • How churches think about participation:
      • Leadership and organizational structure are crucial
      • Emphasis on the smallest reproducible unit in a church, often small groups in megachurches
      • Jesus spent 73% of time with apostles, highlighting relational focus
        • What is the smallest reproducible unit in your church, and how are you investing in that?
        • Small groups as a central strategy in megachurches
    • Reimagining democracy:
      • Critique of “politics as spectacle.”
      • Politics should be a practice, not something given or consumed
      • Start with politics as a practice, a dyad or triad
      • Focus on interests, identities, and preferences
      • Participation should be relational and focus on negotiation
      • Identities are experiential
      • Problems to be solved are about interpretation, commitment, etc.
    • Key lesson—leadership and organization:
      • Leadership and organizational structure matter
      • Small groups as a central strategy in megachurches
    • Politics as spectacle vs. politics as practice:
      • Politics as spectacle:
        • Individualistic “consumer” approach
        • Participation focuses on expression or deliberation
        • Identities are categorical (indivisible)
        • Problems to be solved are about aggregation, representation, etc.
      • Politics as practice:
        • Relational “constituent” approach
        • Participation focuses on negotiation (needs collective action)
        • Identities are experiential (pluralistic)
        • Problems to be solved are about interpretation, commitment, etc.
        • Scale depends on the quality of the tinder and kindling, not just the size of the fuel logs
        • Need to create room for leadership and organization to emerge

    Questions

    • Francesca:
      • Cathy O’Neil’s perspective on near-term artificial intelligence and ethical matrix
      • Consideration of potential stakeholders impacted
      • Well-being, autonomy, and justice framework for all those impacted
      • Hailey’s input on identifying stakeholder groups and giving them units of power reflecting their impact
    • Claudia:
      • Coordination theory by Mary Douglas
      • Need for equilibrium between the need for authority and individual autonomy
    • Collectivities:
      • Recognition of horizontal relationships among members, using credit unions as an example
      • Dimitri’s perspective on possibilities for mobilization in radical democracy models
    • Joshua:
      • Reflection on citizen assemblies and sortition as dominant modes of discourse
      • Exploration of collaboration modes that haven’t been considered
    • Glen:
      • Emphasis on the relevance of leadership within groups
      • Consideration of the importance of accountability
      • Claudia’s input on Lebanon’s consultation as confirming divisions and the need to envision new democratic spaces
    • Aviv:
      • Differentiation between negotiation and deliberation
      • Hahrie’s perspective on the need for both in the real world
      • Leaders as navigators of uncertainty and negotiators who pull people together
      • Importance of having individuals who are masters of politics and not prisoners of ideology
    • Ruthanna:
      • Citizen’s assemblies as a layer among other layers
      • Exploration of subsidiaries and how to include input from the more-than-human community
      • Redefining work in the context of the heritage of Puritan and Calvinist culture
    • Dimitra:
      • Framing of negotiation and deliberation as complementary concepts
      • The need for both in problem-solving
      • Reference to Marshall Gantz’s perspective on leaders showing purpose in the face of uncertainty
      • The concept of the smallest reproducible unit, such as circles of trust

    Bruce Schneier January 8, 2024 7:05 AM

    Panel #8 Getting There

    Lecture

    Nick Chedli Carter: Harvard Kennedy School Ash Center

    • Vision:
      • Aim for a more equal, better functioning society, democracy that is capable to tackle big problems and more responsive
      • Seeding people with a broader range of ideas, building bridges between people, innovation and technology
      • Collaborative innovation towards a better democracy and more responsive government
      • Human-centered, multicultural, bright democratic future
      • Marshall Gantz—Can Democrats be "people-first" if their campaigns value people last?
    • Personal orientation: 
      • Campaigns and electoral politics
      • Issue advocacy
      • Innovations to advance democratic practice
      • Media (from public access to corporate)
    • Critique of current political practice:
      • Online and consolidated
      • Top-down and dehumanizing
      • Macho
      • Isolating and polarizing
      • Dominated by the presidential election cycle
      • Facts diffuse and at odds with democracy
    • Question and opportunity: 
      • How to value and empower local communities – and defining community – in an increasingly online political landscape
      • Learning from the pandemic experience to address significant problems
    • Learning from pandemic experience:
      • Technologies implemented (online voter registration, ballot tracking)
      • Achievements by Congress, state legislators, election officials, and local governments to allocate resources
      • Collaborations between government civil society, faith communities, businesses etc…
    • Ash Center initiatives: 
      • Electoral system reform
      • Election administration profession
      • Technology and democratic practice
      • Parties, associations, and participation
    • Upcoming in 2024: 
      • Local democracy innovation case studies
      • Electoral reform symposiums
      • AI & Democracy Working Group

    Danielle Allen: Harvard Kennedy School Ash Center; Partners In Democracy

    • Personal experience:
      • Ran for governor of Massachusetts in 2020, leading to extensive travel around the state
      • Engaged with local communities, highlighting challenges and perspectives
      • Recognized the disconnect felt by citizens, leading to a sense of not being involved in democracy
    • Three examples of people’s experiences:
      • “It’s just the not knowing”: a person feeling disengaged from democracy
      • Observation of Western Massachusetts feeling overlooked and unrepresented
      • Addressing issues like the absence of beds, emphasizing the need for responsive governance
    • Voting and representation:
      • People who feel unrepresented are less likely to vote
      • Advocates for moving towards a democracy that is responsive and where people feel seen and empowered
    • Learning from political system decrepitude:
      • Recognizes the challenges individuals face when trying to get on the ballot
        • Need to win small delegates at caucuses (only 20% of caucuses qualify)
        • Local news deserts
      • Emphasizes the need for rethinking representation beyond political reform
    • Reconceiving representation:
      • Beyond political reform, focuses on reconceiving representation
      • Aiming for concrete use cases and identifying individuals with the capacity to implement and scale them
    • The power of Cambridge, MA:
      • Historical impact: opened schools in Cambridge, then in the state, and eventually across the country
      • The vision of building digital civic infrastructure in Cambridge and scaling it across Massachusetts and the nation
    • Chance to rethink representation:
      • Advocates for rethinking representation, identifying specific use cases, and collaborating with individuals capable of building scalable solutions
    • Focus on empowerment:
      • Stresses that humans thrive when empowered, highlighting the importance of creating a bedrock of empowerment
    • Specifics on political processes:
      • Describes the need to win delegates at small caucuses to get on the ballot, emphasizing the lack of awareness about this process among many in Massachusetts
      • Addresses issues related to local news deserts and the limited qualification of caucuses
    • Future vision:
      • A vision of building digital civic infrastructure in Cambridge, using it as a model for Massachusetts, and subsequently for the rest of the country

    Marci Harris: POPVOX Foundation

    • Background and phases of technology and democracy:
      • Legal background, worked in Congress
      • Phases: transparency, civic tech movement (open government), social media activism, techlash, pandemic, Gen AI
    • Divide between tech world and DC:
      • Tech world sees rapid, transformative change
      • DC views it as just another hype cycle, slow to change
    • Linear vs. exponential change:
      • Governments operate linearly; tech advances exponentially
      • Congress faces three pacing problems
    • Congress’s pacing problems:
      • External: fails to keep pace with emerging innovations
      • Inter-branch: lags the executive branch
      • Internal: doesn’t employ modern practices and technology for its own operations
    • Progress and challenges:
      • House Select Committee on the Modernization of Congress
      • Modernization accelerated during the pandemic
      • House Digital Service and Modernization Fund
      • GAO Science Tech Assessment, Analytics, and Innovation Lab
      • Staff pay gap addressed, intern resource office, paid internship funding
      • Tech fellowships to bring expertise into government
      • First Branch AI working group
    • Forthcoming paper—“Representative Bodies in the AI Era”:
      • Drawing inspiration from other legislatures
      • Recommendations for phased implementation and oversight
    • Perspectives on the future:
      • Shift from linear to exponential change
      • Perception gap between the tech world and DC
      • Congress’s three pacing problems highlighted
    • Approach to legislative processes:
      • Move beyond digitizing existing systems to accomplishing goals
      • Transition from top-down to bottom-up decision-making
      • Shift from a waterfall-based process to agile, ongoing refinement
      • Decentralize discretion to the local level for participatory processes and deliberation

    Aditi Juneja: Democracy 2076

    • Engagement with political organizing:
      • Only 3.5% of Americans spend time engaged in political organizing
      • Emphasizes the need for politics as a practice
    • Executive orders and border wall:
      • Executive order for building a border wall
      • Challenges the use of executive powers for such actions
      • Won an injunction against the border wall using HEROES Act powers in El Paso
    • Structural reasons for inequality:
      • Explores the tyranny of the minority as a structural reason for inequality
      • Faces resistance within the NGO for discussing this issue
    • Landscape of ideas and national task force on election crises:
      • Created a landscape of ideas to unite organizers and launch a campaign
      • Formed the National Taskforce on Election Crises in 2019
      • Addressing challenges related to the “big lie” after the election
    • Concerns about emergency powers:
      • Concerns about the expansion of emergency executive powers
      • Highlighted Biden’s executive order against canceling student debt
    • Heritage foundation and constitutional amendments:
      • Heritage Foundation supports amending the US Constitution in a likely authoritarian manner
    • Constitution for 2076 and strategic foresight:
      • Gathered organizers to envision a constitution for 2076
      • Explored four future scenarios: continued growth, collapse, transformational, and discipline
      • Considered projections related to demographic change, climate change, and the future of work
    • Proposed constitutional reforms:
      • Proposed constitutional reforms include UBI, wealth tax, right to health, abolishing the Electoral College, rights for education, truth and reconciliation commissions, SCOTUS reform, campaign finance reform, abolishing the Senate, and multi-member districts
    • Democracy as values and rights:
      • Advocates for viewing democracy beyond elections, focusing on values and rights
      • Stresses the importance of practicing politics to bring about change

    Spencer Overton: Multiracial Democracy Project, GWU Law School

    • Background and roles:
      • Election law professor
      • Served in the Obama administration and on the campaign
      • Headed a black think tank until last year
    • Diversity and similar problems worldwide:
      • Similar but different problems in diverse countries like Brazil and South Africa
    • Trends in black state legislators:
      • Immense decrease in black state legislators in the South followed by an increase in recent years
    • Citizenship Act and immigration law:
      • Discussion on the Citizenship Act and its ties to immigration law
    • Demographic changes and political impact:
      • US becoming more diverse, potential impact on political landscape
      • Southern white flight from the Democratic Party
    • Race, voting, and cultural anxiety:
      • Race’s role in party voting
      • Cultural anxiety and its impact on thwarting democracy
    • Multiracial democracy challenges:
      • As the US becomes more diverse, challenges like intersectionality need to be addressed
    • Voting Rights Act:
      • Importance of the Voting Rights Act in the context of multiracial democracy
    • Impact of AI on multiracial democracy:
      • Exploring the effects of AI on the multiracial democracy landscape
    • Digital blackface, racial targeting, and trust:
      • Discussion on digital blackface, racial targeting, and the implications for trust
      • Reference to the “liar’s dividend” and its meaning for trust
    • Racial gerrymandering, voting restrictions, and cyberattacks:
      • Issues related to racial gerrymandering and voting restrictions
      • Mention of targeted cyberattacks
    • Historical perspective:
      • Historical overview of Black state legislators in the South, including the decline after the end of reconstruction
      • Immigration law changes in 1965 that revised the quota system
      • Southern white flight from the Democratic Party between 1960 and 2000 and its impact on racial split in party support
    • Multiracial democracy project at GWU:
      • The GW Multiracial Democracy Project serving as a bridge between academia, policymakers, and other stakeholders
    • Russian targeting and partisan gerrymandering:
      • Russians targeting Black voters and the impact
      • Politicians engaging in partisan, non-racial gerrymandering, considered nonjusticiable

    Rob Ricigliano: The Omidyar Group

    • Background and role:
      • Works with ten organizations in a network founded by Pierre Omidyar
      • Leads the Systems and Complexity team
    • Ecosystem of governance:
      • Started in early 2021 with a framing question: what would make for a healthy system of governance in a fully digital age that supports the well-being of individuals and society?
      • Interviewed fifty-six people from diverse backgrounds and regions, themed into five first principles of health governance and forty associated pathways
      • Conclusion: government and governance are not co-equal concepts; governance is broader than government
    • First principles of healthy governance:
      • Identified five categories: being together, working together, weaving together, caring together, deciding together
      • Being together: building connections and complex understanding of each other, past and future generations, and the environment
      • Working together: increasing agency, both individually and collectively, over the issues that matter
      • Weaving together: understanding and connecting identities, worldviews, and stories to heal and live in light of them
      • Caring together: expressing and enacting care for the needs of individuals, communities, and the environment
      • Deciding together: collectively stewarding communities and the environment, setting and enforcing rules
    • Building a healthy governance ecosystem:
      • Requires a change in the dominant paradigms of democracy
      • Governance Future Network formed by many interviewees and Omidyar Group members
      • Focus on transforming how we collectively steward the interdependent well-being of people, planet, and technology
      • Emphasizes four critical shifts and cross-network learning, four current initiatives
    • Current initiatives:
      • Ritual Studio
      • Xcavate: collective futures generator
      • Town Hall 3.0
      • Democracy Maker Space

    Questions

    • Sam’s questions:
      • How many fractures in democracy are problems of recognition?
      • For all the innovations addressed, which ones reflect problems of recognition?
      • Spencer’s insight on the problem with white identity politics and multimember districts
      • Danielle’s emphasis on linking recognition and efficacy in engagement
    • Claudia’s perspective:
      • Lack of investment in developing democracy beyond elections
      • Aditi’s point about people dropping out and the need for governance to be of the people
      • Nick’s idea about collaboration opportunities during the census
    • Rob’s insights:
      • Recognition of pieces of identity
      • The two-step dance: recognition and giving agency to recognized individuals
      • Changing the philanthropic architecture for systemic changes
    • Dimitri’s question:
      • How do we get there? The role of conjunction of actors
      • The importance of a conjunction of actors for progress
    • Aditi’s reflections:
      • Organizers focusing on structural reform rather than offensive work
      • The need for different stories and narratives shaping possibilities
      • The simultaneous need for changes in the narrative and political environments
    • Helene’s concerns:
      • The generational gap and obstacles in technology
      • Contrasting views between those in power and new generations
      • Marci’s optimism about opportunities for change and the impact of data
    • Riley’s perspective:
      • Feeling of not being seen and questioning institutions
      • Spencer’s response emphasizing constructive nonviolence
      • Rob’s focus on offense vs. defense and building a different house
    • Bruce’s observation:
      • Ideas lying around and policymakers looking for them during crises
      • Identifying the two most unrepresentative groups in Congress
    • Additional questions from set 2:
      • Problems of recognition?
      • Disillusionment with politics
      • Can we reimagine a democracy into something more effective?
      • MAGA out-organized centrist Republicans
      • Why put energy into this system that hasn’t served us in the first place?
        • Windows of opportunity that come along so the debate needs to happen now before the window comes so you can take advantage of them
        • Nonviolent resistance is not necessarily the master’s tools
      • AI and the changes could be one of those moments to make change
        • Example: tornado hits town right before election and now you have to recover; status quo not an option
        • Data and AI will change things; systems are going to be more effective soon
        • AI can be used to make ESL learning plan (use it to translate reports so families who don’t speak English can understand them better)
      • Organizers wish they were doing more offensive work and don’t think longer than a five-year timeframe
        • Organizers are in a defensive, reactive posture and can’t think about the future
        • Have to imagine new and audacious futures
        • Organizing tied with storytelling—narrative environment is very doomer
        • Parties realign every thirty years and we’re living through one right now
          • What seems impossible changes when realignment happens
      • Notion of ideas lying around—when there is a crisis, policymakers look for them. They’re formed when they’re needed
      • Two most underrepresented groups in Congress: non-college-educated people, blue-collar workers

    Clive Robinson January 8, 2024 8:03 AM

    @ Bruce, ALL,

    Re : Evolution is continuous change by definition, otherwise it would be stagnation.

    “Everyone seems to be talking about ways to reform our existing systems. That’s critically important, but it’s also myopic. It represents a hill-climbing strategy of continuous improvements. We also need to think about discontinuous changes that you can’t easily get to from here; otherwise, we’ll be forever stuck at local maxima.”

    A hill-climb implies only improvment that is going upwards. Reality is climbing hills is like the swinging of a pendulum it goes up and down and averages out to some mean value.

    In reality politics is a response to society and society changes continuously.

    The aim of the game is for politics to in some way to reflect some part of society but by design it has to move in a stepwise fashion as legislation and regulation almost invariably lag behind the mores and morals of society.

    Historians tend to see the changes in society as the primary measure of their domain of expertise over the long term. However most people only see the gap between society and politics as it changes in the short term. It’s a quite important difference in point of view.

    Thus whilst historians see a progression of evolution towards some destination that although it continuously changes slightly it tends toward a given direction (often for the better). What the majority sees though appears to be the almost random back and forward swing of the gap, thus see a pendulum like “drunkards walk” rather than progress.

    At the moment it would appear that politics is moving in a negative direction away from the societal mean. Which raises the question of what is likely to happen.

    History suggests that through either changes due to pandemic disease or significant civil unrest society will lessen the gap and even reverse it’s direction.

    However some have found that exploiting fermenting civil unrest with the political structure can be redirected away from the actual cause to another thus apparently absolving the political structure.

    However as has been observed and notably highlighted by George Orwell there is only so far the absolving can go before it gets pushed from potential insurrection of civil conflict to actual interstate conflict and large scale warfare.

    Some have observerd we are currently on a cusp or tipping point where large scale if not global warfare can become a certainty.

    Thus the important questions are,

    1, How might we step back from the edge?
    2, Can we make the changes in time?
    3, If not how do we limit the harm?
    4, How do we clear up the mess?

    In previous posts I’ve made one or two suggestions for apparently quite minor changes.

    Firstly taking the money/profit out of politics.

    Secondly having a “sun set clause” in all legislation.

    Thirdly taking out “veto clauses” in any political and legislative system.

    The third I’ve specifically pointed out in the past with regards the United Nations and the Security Council. It can only lead to international conflict in the long term, and that is exactly what we are currently seeing happen.

    Aaron January 8, 2024 10:39 AM

    I’m glad somebody with a rational mind is thinking about the future of Democracy. It seems, regardless of the political party in charge, they are losing a grip on what Democracy actually means.

    Bob January 8, 2024 11:37 AM

    Otherwise rational people pretending this is a “both sides” issue adds as much harm to the situation as anything.

    Clive Robinson January 8, 2024 11:39 AM

    @ Aaron,

    “they are losing a grip on what Democracy actually means.”

    The reason for this is you do not live in a Democracy, at best it’s a “Representational Democracy” as part of a Republic or similar.

    That is you don’t get to decide on anything of concern very often if at all.

    Think of it more like a “Mr Owniverse” or similar beauty pageant. There are a bunch of people who you’ve in no way actually selected, that you have to then chose one of based on what is usually a bunch of lies.

    Once they have been chosen you the voter have no control over them what so ever, but those with large amounts of money do.

    The thing is appart from the large amounts of money mostly they are not there for you but them. That is their narcissism, sadism, and other highly unpleasent traits found in psychopaths etc.

    These are not people you would like comming into your home so why would you chose one?

    Because you have no other choice. The late author Douglas Adams had a joke about the whole process,

    Ford Prefect looked at Arthur Dent and said “On its world, the people are people. The leaders are lizards. The people hate the lizards and the lizards rule the people.”

    “Odd,” said Arthur. “I thought you said it was a democracy.”

    “I did,” said Ford. “It is.”

    “So,” said Arthur, hoping he wasn’t sounding ridiculously obtuse, “why don’t the people get rid of the lizards?”

    “It honestly doesn’t occur to them,” said Ford. “They’ve all got the vote, so they all pretty much assume that the government they voted in more or less approximates to the government they want.”

    “You mean they actually vote for the lizards?” asked Arthur with incredulity.

    “Oh yes,” said Ford with a shrug, “of course.”

    “But,” said Arthur, going in for the big one again, “why?”

    “Because if they didn’t vote for a lizard,” said Ford, “the wrong lizard might get in.”

    One of the things I want to change on Balot Papers, is to have included as the last option,

    [ ] None of the above.

    But apparently for some reason that is not democratic…

    Winter January 8, 2024 12:02 PM

    @Aaron, Bob

    Only one side is actively dismantling democracy.

    I just heard one side was selling out America to the Chinese.

    Guess which side:
    ‘https://amp.cnn.com/cnn/2024/01/04/politics/trump-properties-china-foreign-payments/index.html

    ‘https://amp.theguardian.com/us-news/2024/jan/04/trump-businesses-payment-house-investigation-china-saudi-arabia

    Bob January 8, 2024 12:03 PM

    Reagan swept the country and mortgaged the future to give his voters peanuts in the moment. By the time Clinton came in, he was effectively running on right-wing economics with barely left-of-center social policies.

    The American voter pushed the Overton window to the far right. The American voter mortgaged the future for peanuts in the moment, and liked those peanuts so much that they re-elected the guy by an even wider margin. Now that the bill for that mortgaging is due, I’m not going to pretend the mortgage wasn’t taken out to the sound of thunderous applause by the American electorate.

    Winter January 8, 2024 12:09 PM

    @Clive

    at best it’s a “Representational Democracy” as part of a Republic or similar.

    It is an Oligopoly where various billionaires try to take over the “ruling” parties, or their candidates. That is especially strongly present in the US as bribing politicians is legal in the US. It is much less legal in many other countries.

    Winter January 8, 2024 12:12 PM

    @Bob

    Every GOP accusation is a confession.

    Indeed, that is an area where there is little room between Putin and the GOP: Every accusation is indeed a confession.

    Bob January 8, 2024 12:28 PM

    @Winter

    That’s because they’re the same picture. Which is why seeing these false equivalences from people I consider intelligent and rational is both frustrating and terrifying.

    lurker January 8, 2024 12:58 PM

    Start with the best one-liner:

    Robert Jordan Hall: It is by no means obvious that the epistemology of the west is true or obligatory.

    Although that could be difficult for some. In some parts there is a widespread belief that any suggestion of anything of the west being untrue or not obligatory, is itself “undemocratic.”

    Bob January 8, 2024 1:19 PM

    @lurker

    Show me who’s doing democracy better outside of the West. Looking somewhere different just because it’s different is just a hair’s breadth from the politician’s syllogism.

    At a certain point, it feels like people are throwing shit at the wall to see what sticks. I assume that’s because it’s more comfortable than acknowledging the fact that a full third of the US voting population, as well as one of our two major political parties, are overtly fascistic.

    Erdem Memisyazici January 8, 2024 1:49 PM

    Cory Doctorow was spot on last time. I liked his blackhole analogy where rules similarly seem to break down in institutional structures when power is concentrated.

    You really have to experience corruption first hand to see it.

    My personal experience has been with being drugged and interrogated in a frat house then injected with speaker implants to help repair the neurological damage. These people actually contacted a real doc

    Erdem Memisyazici January 8, 2024 2:00 PM

    I liked what Cory Doctorow had to say last time. He was spot on with his analogy on blackholes and concentrated institutional power structures where similarly rules seem to break down. You have to experience corruption first hand to see it.

    My personal experience involves being drugged and interrogated about my personal beliefs on Edward Snowden etc. in a frat house where the people who did so actually contacted a real doctor to inject my drugged self with an implant that supposedly repairs the neurological damage received from your will being broken as you are compelled by powerful psychoactive drugs to reveal your inner most secrets.

    The same institution (the one I happen to still work for) is now in a position to either be truthful or somehow explain why they were assisting individuals with behavioural problems surely they had before they got done with them in a frat house.

    It took me 10 years and more money than I have to find out I wasn’t just followed everywhere by frat members but that this is indeed a legitimate medical procedure that is performed nationwide for anyone suffering from post-traumatic stress.

    Personally I am crossing fingers people simply speak the truth in court, but only time will tell.

    Much like a blackhole even the nicest people in society like doctors can be swayed and partake in corruption that is just too evil to fathom if everyone in power is in absolute agreement and stand to benefit in one way or another.

    Anon E. Moose January 8, 2024 2:32 PM

    A democracy is still the worst possible form of government, except for all of the others. This is why, in the United States, we do not have a democracy. We have a Constitutional Representative Republic. Which is a blessing. None of you have the ability or time to understand all of the changes that are made on a daily basis in our country. Barely can most of us understand the spoon-fed propaganda from one side of the media or the other.

    When this form of “future” democracy is attempted we will see a new form of graft, fraud and the vote buying will be second to none. Why? Zero moral values exist for to many people.

    Aaron January 8, 2024 2:58 PM

    @Clive

    “[ ] None of the above”
    It’s a good idea but why not allow for -1 votes? It’s either you have a +1 or a -1 vote, not both.
    If you don’t like the candidates, then you vote against them rather then voting for “the lesser of two evils”

    Douglas Adams, through Ford, nails the perspective of human governance. The reference to lizards is perfect too if you consider our more primal brain functions are essentially that of a lizard or reptilian evolutionary lineage. Regardless of our higher cortex functions, the hind ‘lizard’ brain can, at a whim, threaten the whole system until it gets what it wants.

    @Winter

    You’re viewing range is too narrow, expand it out a century or so.
    It’s both sides, they just roll reverse when their is a majority ownership of offices.
    Why do your elected job when you can spend your time undoing what the last elected individual did all while propping up defenses against next person who will go after what you’ve put in place.
    https://www.youtube.com/watch?v=tEczkhfLwqM

    Every political accusation is a confession, for every politicians is on a journey to mastering projection.

    Bob January 8, 2024 4:18 PM

    @Aaron

    “Both sides” serves only to elevate the worst. It’s an intentional muddying of the waters by those who sympathize with the worst.

    Clive Robinson January 8, 2024 5:31 PM

    @ Bob, lurker,

    “Show me who’s doing democracy better outside of the West.”

    Look at the “bushmen” in the Kalahari Desert, or nearly all supposadly “primitive cultures” where all in a small group reach a consensus on any given action of import to them.

    The problems with Democracy are,

    1, It does not scale
    2, above a certain size hierarchies form.
    3, Hierarchies focus power upwards.
    4, Power attracts certain undesirable people and behaviours.

    Sadly most people either,

    1, Feel their voice does not count.
    2, Do not want the responsibility.
    3, Have narrow agendas
    4, Have self interests they see as rights.

    It’s why I point out from time to time,

    “Individual Rights v. Social Responsibility”

    Thus,

    “Without open oversight all hierarchical structurs become corrupt.”

    Thus unless we solve the issues of,

    1, People taking responsability.
    2, Implicit failings of hierarchical systems.

    What we call “Democracy” will be at best a sham used to hide the actions of those with ill intent with regards to society.

    Why this is not more obvious to people actually surprises me, even though I know it should not.

    Just remember to very scary sayings,

    1, “For the common good.”
    2, “A safe pair of hands.”

    Neither is even remotely true.

    lurker January 8, 2024 5:54 PM

    @Bob

    Show me who’s doing [Western] democracy better outside of the West.

    There, fixed that for you. As @Anon E. Moose said, you don’t have a democracy in the US, because … moral values. Whose morals? whose values?

    Clive Robinson January 8, 2024 6:02 PM

    @ Aaron,

    ” but why not allow for -1 votes? It’s either you have a +1 or a -1 vote, not both.
    If you don’t like the candidates, then you vote against them rather then voting for “the lesser of two evils””

    Two related reasons,

    Firstly as you analyze it you will find it becomes just another form of “failing proportional representation”.

    Secondly it alows the “No Confidence” vote to be not just hidden but unacted upon.

    I know it sounds odd but I realy want rid of “representational democracy” because it clearly is not “Democracy” and worse it positively encorages corruption in nearly all it’s worst forms.

    Have a chat with someone who has taken a long considered look at “Scotty from Marketting” who was extrodinarily corrupt even by the “Corrupt standards” of Australian politics.

    I can make similar comments about India, Indonesia, US, UK and many many more.

    Just remember where ever you see the expression,

    “liberal democracies”

    What is in reality being discussed is an asocial very probably corrupt system, that those who benifit by it personally will fight to keep at the expense of society (including themselves at the end of the day).

    Bob January 8, 2024 8:06 PM

    @lurker

    You’d prefer to look at Eastern Democracies? South American? African? Clive came up with something. Granted, it was something we were trying that proved unsustainable at OWS rallies however many years ago that was, but still. It’s something other than “u is brainwashed westerners” at least.

    &ers January 8, 2024 9:30 PM

    @ALL

    Speaking about “democracy” :

    hxxps://cdn.statcdn.com/Infographic/images/normal/18737.jpeg
    hxxps://www.visualcapitalist.com/mapped-the-state-of-global-democracy-2022/

    Yes, there’s a lot of the flawed kind of it around…

    Winter January 9, 2024 1:59 AM

    @Aaron

    It’s both sides, they just roll reverse when their is a majority ownership of offices.

    That is part of the corruption of the US two party system. Where a president denies patients life saving respirators because they live in a state governed by the other party.

    ‘https://www.npr.org/sections/coronavirus-live-updates/2020/03/27/822998602/fact-check-n-y-governor-slams-trump-ventilator-claim-as-ignorant-and-uninformed

    Bownse January 9, 2024 9:38 AM

    Democracy is 2 wolves and 1 sheep voting on what’s for lunch. It’s a good thing that the USA was never designed to be anything other than a Representative Republic. It avoids “mob rule”.

    Bob January 9, 2024 9:42 AM

    @Bownse

    Yeah. Much better what we have now, where a single wolf outvotes half a dozen sheep in deciding what’s for dinner.

    Keith Douglas January 9, 2024 9:47 AM

    What a wonderful sounding conference. Thanks for sharing.

    On the methods of voting question some are discussing: as a semi-joke, in high school some of my friends and I ran for student’s council executive with the party name “None of the Above”. We were one draw of the lottery away from actually being last on the ballot, too! (No, we did not win.)

    Kent Brockman January 9, 2024 1:53 PM

    What makes our “undemocratic democracy” here in the US even worse are the ‘unelected nine’ who serve life terms and who are deferred to despite, as a body, being equally if not even more corrupt than the other branches of govt. while its legitimacy is seemingly degraded by the day. The twin stupidities of the Electoral College and life terms for SCOTUS judges haunt this country and make a mockery of even our pretend democracy.

    Bruce Schneier January 9, 2024 4:58 PM

    @Bownse:

    “Democracy is 2 wolves and 1 sheep voting on what’s for lunch. It’s a good thing that the USA was never designed to be anything other than a Representative Republic. It avoids ‘mob rule.'”

    It’s not the republic nature that avoids that problem.

    One of the most important things to determine in any democracy — or republic, or whatever you want to call it — is what is off-limits to democracy. Can we vote ourselves all the money? Can we vote you in jail? Can we vote to forcibly take away your kidney? And on and on and on.

    All democracies have rules about what isn’t allowed to be voted on.

    lurker January 9, 2024 6:43 PM

    @Kent Brockman
    re, ‘unelected nine’

    They are “elected”. POTUS and Senate elect them, similar to the way members of the Politburo are elected, except Politburo members can be un-elected if they don’t toe the Party line of the day.

    Clive Robinson January 9, 2024 8:26 PM

    @ Bruce, ALL,

    “One of the most important things to determine in any democracy — or republic, or whatever you want to call it — is what is off-limits to democracy.”

    Actually that is to do with “representation” not “democracy” and is fundementally a trust issue.

    When a group is small, all can participate more or less equally thus trust tends not to be an issue and it does not need to be formalised as such.

    As a group gets larger it’s nolonger practical for all people to participate equally. Initially it starts by sub-grouping where groups of similar point of view select individuals as spokes people to go off to other groups. This almost always falls into a hierarchy with selected individuals from sub groups form super-groups where issues are most likely resolved.

    The problem is in all hierarchical systems power moves upwards and those at the bottom start to need “formalized trust” to protect them from betrayal etc. The larger the base layer size the larger the number of intermediates are needed and where power is concerned trust should be thin on the ground and the negotiation power of representatives thus formally limited.

    Kent Brockman January 9, 2024 8:30 PM

    @ lurker

    That is a form of “election” if you like, however, being a third branch of government they should have been put before the people as the member(s) of the other two are. More aggregious are the lifetime appointments where an appointee could conceivably sit for 40-50 years with virtually zero chance of recall.

    Just Passin’ Thru January 10, 2024 1:10 AM

    Whoa!

    That’s some response!

    My only comment is that the re-imagining of democratic institutions, based on contemporary technology, should be robust enough to withstand assaults on that technology. For example, I’ve read that the detonation on 4 EMP weapons above the continental US would render our electronic-based technology useless. (You might have a blockchain-based scheme, but without the infrastructure it would be less trustworthy or trusted than even our current deep state.)

    Winter January 10, 2024 1:25 AM

    @Bob

    Democracy is 2 wolves and 1 sheep voting on what’s for lunch.

    That has it backwards.

    There are, and will always be, more sheep than wolves [1]. So it is never 2 wolves for every sheep. Therefore, Democracy is 2 sheep and 1 wolf voting on what is for lunch.

    But that sounds rather too pro-democracy for many.

    [1] This is basic ecology and biology. But I understand that the American Billionaires think of themselves as poor sheep that are under siege of all the people they extracted or extorted their money from. But it is rather telling that Americans think of Jeff Bezos as a sheep and his warehouse “employees” as wolves.

    Winter January 10, 2024 2:55 AM

    Democracy starts and ends with freedom to get informed about politics, society, and life:

    Lately, Americans love to ban books about society and life’s choices, at least some Americans.

    Actually, Just 11 people were responsible for most 2021-2022 school book challenges. A Virginia woman challenged 71 out of the 73 books she read.
    ‘https://www.businessinsider.com/school-book-challenges-bans-virginia-moms-liberty-education-2023-9

    According to an analysis by the Post, 60% of book challenges made in the 2021-2022 school year came from the same 11 adults. Petersen was one of them. The majority of objections were on books authored by or about LGBTQ+ people or people of color.

    Clive Robinson January 10, 2024 8:54 AM

    @ Winter, ALL,

    Re : Lunch and who / what.

    “Therefore, Democracy is 2 sheep and 1 wolf voting on what is for lunch.

    Actually no, not quite it’s

    “2 Sheep voting on who is going to be the Wolf’s lunch”

    That’s why it’s a matter of trust as @Bruce notes above.

    However the history of societies shows that where the “Hawk-Dove” issue exists, there is a very important fact,

    “If you don’t have a say then you become the prey.”

    As can clearly be seen by outsiders to the US System, and some US insiders.

    But as for the other insiders they either don’t see or don’t care so become prey for the slaughter. Which makes the US system quite deliberately the modern equivalent of the Ancient Roman Circuses[1] only the “Prolls”(Proletariat) don’t get any free bread. With some “patricians”(ruling classes) want walls and barriers to keep the “Plebs”(plebeians) out or voteless.

    Oh as for Circuses they were in the time of no 24hr TV in effect the only way for politicians to bribe the “Prolls” to vote for them,

    https://en.m.wikipedia.org/wiki/Circus_Maximus

    As oft observed,

    “The more things change, the more they stay the same.”

    [1] Ever wonder why politicians and the like don’t say what it realy is ie “dog eat dog”, but some numpty expression such as “the full circle of life”? Remember that the root word for both Circle and Circus are the same…

    PaulBart January 10, 2024 9:02 AM

    @Clive
    “Without open oversight all hierarchical structures become corrupt.”

    An empire can not have open oversight. And what the world has currently are many empires.

    @Bob
    “one of our two major political parties, are overtly fascistic.”
    And the other is overtly Marxist. I <3 over-generalizations.

    "American Billionaires think of themselves as poor sheep"
    And what of the petty bureaucrats. How do they fancy themselves?

    And remember kids, only one country is allowed to be called "democratic", yet it consists of one people, and one religion, with closed borders, and can never be criticized. Only those in other democracies that advocate for their people, and their religion, and their borders be closed are to be ostracized and called fascist.

    Winter January 10, 2024 9:18 AM

    @PaulBart

    “American Billionaires think of themselves as poor sheep”
    And what of the petty bureaucrats. How do they fancy themselves?

    Who cares about employees?

    Bureaucrats are simply employees who get a salary to do things.

    As much as Americans (and others) are wont to complain about them, whenever these employees cannot or will not work, people wallow in agony. [1]

    Billionaires, however, are people who get rich from employees who get poor, or die as those poor ones working in Amazon warehouses, or the countries slaughterhouses. [2]

    [1] No police, or no fire brigade, or no teachers, or no courts, no public transport, or no…

    [2] ‘https://sentientmedia.org/covid-19-hit-slaughterhouse-workers-harder-than-we-thought/

    Winter January 10, 2024 10:10 AM

    @PaulBart

    @Bob
    “one of our two major political parties, are overtly fascistic.”
    And the other is overtly Marxist. I <3 over-generalizations.

    Not quite “over-generalizations”.

    Lets see what the prefered candidate of the “one” party says about his political enemies:

    “We pledge to you that we will root out the communists, Marxists, fascists, and the radical left thugs that live like vermin within the confines of our country,” he told a New Hampshire crowd. [1]

    and about immigrants:

    “They’re poisoning the blood of our country. That’s what they’ve done,” Trump told the crowd.

    “They poison mental institutions and prisons all over the world, not just in South America … but all over the world. [2]

    Now, we have these same words from other politicians that did self-identify as (and were proud to be) fascists:

    “I will get rid of the ‘communist’ ‘vermin’,” “I will take care of the ‘enemy within’,” “JJJJ and migrants are poisoning Aryan blood,” and “One people, one realm, one leader.” [3]

    So, we have one party whose most powerful politician openly sprouts fascist propaganda.

    Now, what “Marxist” propaganda did we receive from the other party?

    [1] ‘https://www.npr.org/2023/11/17/1213746885/trump-vermin-hitler-immigration-authoritarian-republican-primary

    [2] ‘https://www.theguardian.com/us-news/2023/dec/16/trump-immigrants-new-hampshire-rally

    [3] ‘https://www.newsweek.com/fact-check-donald-trump-adolf-hitler-viral-quote-comparison-accurate-1843501

    Bob January 10, 2024 1:38 PM

    @Winter

    PaulBlart isn’t going to respond to you, because his only goal here was to muddy the waters with fascist talking points. If it weren’t for false equivalences, the right wing would have no equivalences at all.

    Winter January 10, 2024 1:54 PM

    @Bob

    PaulBart isn’t going to respond to you,

    I know. My responses are partly addressed to the comment, but mostly for other readers (and search engines).

    Falsehoods should always be exposed, preferably directly next to the falsehoods.

    Furthermore, I choose to interpret non-responses as admitting “defeat”. [1]

    [1] There is actually no defeat in discussions as discussions are not fights. There is eventually just a lack of counter-arguments or interest.

    Bob January 10, 2024 2:46 PM

    @Winter

    Yeah, I’m on the same page. It’s just easier for me to add context by speaking to you. Personally, I appreciate you pushing back on the propaganda. I don’t want to have to be the one to talk to these “people.”

    That said, I do actually differ in my view of these discussions. The propaganda stage of WWIII is already well underway, happening all over the internet in places and discussions just like this. The first casualty when war comes is truth. The battle for the truth is here and now.

    Clive Robinson January 10, 2024 3:39 PM

    @ Bob, Winter, ALL,

    Re : Three sign posts of disaster.

    “The first casualty when war comes is truth.”

    According to the “three sign posts” it’s actually the second not the first…

    The sign posts are,

    The first only becomes recognised with restrospect / hindsight.

    The second is only recognised by a few at the time but becomes obvious with the third sign post or when the disaster has started/happened.

    The third is widely recognised but for various reasons gets ignored so even though avoidable the path is set for full steam to disaster.

    As a guess I would point at two things as being the first sign post,

    1, Rise of idiot “Strong Men”.
    2, Rise of retrograde religion.

    Both appear to go hand in hand but even they had “enablers” or “predictors” if you prefere.

    Based on that I would say “short yerm neo-con thinking” is very probably one enabler as it gave rise to the idiocy of,

    1, Off Shoring.
    2, Out sourcing abroad.
    3, Loss of supply chain control.

    All three are as spears in the side for an economy and few will survive or recover as skills and market are lost.

    I’ve been soundong off about the fact these three are certainly steps to disaster since the 1990’s.

    At near a third of a century ago,w e’ve had plenty of time to rectify the situation but we’ve not. In fact certain politicians have openly encoraged the stupidity as have certain schools of economic thinking.

    @Bob I don’t have a feeling for how old you are, but I suspect both @Winter and I will not be around come the end of WWIII if we alow it to happen. But with the rise of idiot Strong Men and Religion that forms a very toxic mix as seen with Russia I realy can not see enough people going against the flow to pull us back from the tipping point where Global War starts and goes on for a half decade or more.

    Since the 1990’s the two areas that have concerned me are the Middle East with conflict now there that appears to have been obvious with reyrospect. The other is the south China Seas going into Indonesia, down to Austrilasia and along the Pacific West. Although technically not at war there yet, I think few would argue it’s not a very real hot spot.

    Less obvious but could be the proverbial spark in the powder room, is the Strong Man idiocy in three or more South American Nations. Essentially over the control of resources the West wants cheaply to maintain it’s overly high standard of living compared to the rest of the globe.

    Over forty years ago I was involved with conflict originating there, but I did not see it potentially becoming a global conflict point.

    However over less than a decade things have changed significantly in South America, and it would take very little indeed for it to become the “Ground Zero” of WWIII. Likewise India is rather more than rattling the saber on three sides and has upset three super powers, and shows little or no intention of stopping.

    Which kind of left Africa untill what is happening at the East of the Mediterranean and is now spreading down the East Coast of Africa as well as recently starting along the top North coast.

    Egypt and Turkey are now two points where conflict starting in a major way now appears likely.

    A look in an atlas suggests we are all in range of nuclear weapons from one or more places and the incentive not to use them appears to lessen dramatically on a day by day basis.

    So glowing in the dark appears is an ever increasing probability.

    Clive Robinson January 11, 2024 10:18 AM

    @ Bob,

    “Odds are I’ll be in the mix come whatever may.”

    Well if you can get out of line, via protected occupation etc. Or if you do get selected, to quote another on such things, try to,

    “Keep your head down lower than a snakes belly in a wheel rut.”

    I was once told the two secrets to getting old were,

    1, Stay healthy, happy, and spry.
    2, Never volunteer or be volunteered for anything, as there’s always a catch when they won’t do what they are calling for volunteers to do.

    Bob January 11, 2024 10:29 AM

    @Clive

    Step 1 for me is actually staying out of the box cars if the fascists make enough progress to start overtly rounding up queers.

    JonKnowsNothing January 11, 2024 1:44 PM

    @Bob, @Clive, All

    re: actually staying out of the box cars

    If you have ever experienced a “Command Voice” you will know that’s very unlikely.

    Civilians, rarely come into range of a Command Voice, other than the poorly executed ones from over agitated, highly nervous, police officers giving contradictory commands or commands that are impossible to execute (mutually exclusive directions).

    Animal trainers, successful ones, learn Command Voice and Command Presence. Their voice and body posture aligned with the type of animal they are working with and appropriate to the species behavior.

    Sometimes known as a Sargent’s Voice or Sargent-Major Voice, the effect is immediate compliance. It is a learned method for getting a desired reaction. Some people can do this naturally, others have to stand on the drill grounds and shout until it clicks.

    Abusive and controlling people learn to do it early in life and use it to intimidate others. It taps into our normal fears of harm. Humans are animals and we respond with similar reactions. The sound is not necessarily loud, it can be issued quietly while obtaining the desired effect. It is not necessarily a menacing growl either.

    Confronted with a Command Voice

    • You will get into the box car if told to do so.

    iirc(badly)
    A documentary about WW1 survivors, a man was asked why did he walk into the active fire of machine guns. His reply was a mix of “The officer’s ordered us over the top” (Command Voice) and “All I thought about was going 1 step farther” (Terror).

    ===

    Search Terms: Military Command Voice

    • The Command Voice
      A correctly delivered command will be understood by everyone in the unit. Correct commands have a tone, cadence, and snap that demand willing, correct, and immediate response.

    Bob January 11, 2024 2:30 PM

    @JonKnowsNothing

    At the population level, maybe. But presuming population-level trends upon any given individual or enclave is just bad practice.

    PaulBart January 11, 2024 2:53 PM

    @ALL
    “pushing back on the propaganda”

    What propaganda, that FBI was working with twitter to silence non-Marxists?

    “and about immigrants”
    Ya mean unvetted illegals?
    So the disaster at the border is welcomed by y’all. I hope y’all be opening up your checkbooks and Venmo to pay for this fiasco. Oh wait, no no, socialists and communists only want neighbors to pay for their bleeding hearts. They love stealing from others, their families, their friends, and the future of a nations childrens’ to pay for their messes.

    I know Mexico and their upright police, Columbia and their cartels, El Salvador and their rape gangs, bewwwwtifull places.

    Posting:
    I post just so those reading can do the research themselves. Just making sure your disgusting Marxist lies get exposed as such with facts I post.

    I just post facts.

    Bob January 11, 2024 4:15 PM

    @PaulBlart

    “Non-Marxist” applies to such an overwhelming majority of the US as to be entirely meaningless. You’re using the juxtaposition to try to paint the FBI and/or Democrats as “Marxist.” It’s just another attempt to muddy the waters.

    JonKnowsNothing January 11, 2024 4:47 PM

    @Bob, @Clive, All

    re: basing population-level trends upon any given individual

    @Clive’s recommendations are excellent and he has many posts about surviving group culls.

    There are well documented behaviors of such survival tactics. They all amount to the same thing.

    • Be a zebra
    • Stay inside the herd
    • Do not hang on the outside edges
    • Avoid the splash (1)

    Lions will notice

    ===

    1) Splash the Herd is a ranch and horse competition term meaning to separate one or more cows from a herd. Some competitions your move your horse into the herd without causing the rest of them to panic. In others the more panic you can create when you splash the herd the easier it is to target a selected animal.

    Predators that do not rely solely on ambush, use this method to carve out their selected dinner item from the rest of the menu.

    lurker January 11, 2024 5:07 PM

    @JonKnowsNothing

    So, to run a successful democracy, you also need the Voice.

    ‘https://en.wikipedia.org/wiki/Bene_Gesserit#The_Voice

    Clive Robinson January 12, 2024 4:34 AM

    @ Winter,

    Re : Narco War.

    “Narco states paid for by the American war on drugs.”

    It needs to be said that the US is not the only nation invested in this idea.

    Aside from the anti-XXX advantages you mention it’s a massive boon-dongle of cost to society via a pointless empire of authoritarians whose incomes are very much dependent on creating the story.

    I used to be involved with Pirate Radio and a continuous drum bang from various “guard labour agencies” run by or for the Government was “Drugs” / “Drug Violence” / “International Drug and other Crime”.

    It was as far as I can tell compleate nonsense without any evidence. So we started chalenging it in various ways.

    Interestingly it was after that, that there were occasionaly “new entrants” into the game. Nobody knew who they were which was odd in a very close knit community.

    But then there were raids of these unknown stations where supposadly weapons and drugs in significant quantities were found and significant violence encountered. But strangely even though there were head lines there were no coresponding arrests, suspects or convictions… So “Ghost Operations”?

    Importantly some of us went investigating these Ghosts, and what we found was interesting… “equipment” supposadly taken as evidence by OfCom and “destroyed” was turning up on their transmitter sites.

    Some of the equipment had been deliberately “modified” in various ways to make the output unclean and thus cause more complaints from “aircraft” and the like of “harmful interferance” to navigation and communications.

    Then we started hearing rumors about what was going on in OfCom from insiders.

    Apparently a small team of three people under the leadership of Clive Corrie[1] were running the equivalent of a black operation. Setting up Ghost stations to be raided, we assumed because they had been a total failure against the actual Pirates and needed a few successes to keep their jobs.

    So it was decided to do OfCom’s job against the Ghosts first. A couple of people who knew what they were doing went and turned off and removed the TX site equipment. This was “hidden away” to stop it getting put back into either OfCom or the Ghosts. At first the Ghosts continued, but then I guess OfCom were running out of equipment to use. As almost over night old equipment allegedly stolen from BBC remote sites started turning up, even though no one could say when it had been stolen.

    Eventually the Ghosts did start raising the stakes as gangs of thugs that were at first unknown started turning up. They also started doing significant damage to Council Tower blocks and so the game got escalated and certain Met Police “territorial” groups got involved and turned up with the sole intent to “crack skulls”.

    It thus became clear that part of the plan was to “escalate to violence” by OfCom so that they could push the “Politically Inspired” stories again.

    By this time most of the “old hands” had got out of the game, because the thugs not getting the desired effect started going after the “engineers” and their families including young children. We never did find out at the time if the Thugs came from OfCom or the Met Police. But we noticed that the Met were certainly involved as putting an emergancy call in made the Thugs disapear compleatly before “the local police” turned up when there was nothing to see.

    So a counter-surveilance operation on the Thugs was started… By putting trackers on their vehicles and “mapping them out” as well as filming and photographing them.

    In one or two of the photos it was clear that they were using “official style” hand held radios and as I found out they were using UK “Home Office” frequencies and we made recordings. Further investigation via credit checking agencies showed that the names they were using at the addresses were effectively false as they had no “History or back stories”.

    So much to quite a few peoples surprises motor-cycle couriers turned up at those addresses when the Thugs were out and got the “women” who did have histories and back stories to “sign for delivery” (and in the process get their finger prints).

    Later deliveries were done that had photos and audio tapes and a few other documents in them… And surprise surprise the Thugs melted away.

    Some time later as we all now know the Met Police were caught running such operations and were taken to court for amoungst other things “child support” because their undercover operatives had been fathering children with the unsuspecting women they were hidding behind…

    A few years passed and the old hands were well out of the game and in several cases had started entirely legitimate businesses that were out compeating older businesses. It was fairly easy to outcompeate them as their markups were ludicrous and their designs old and frankly to fragile to be put on the market. But they had contacts in government…

    Clive Corrie came knocking against some of those old hands who had in the past manufactured equipment for pirates. It quickly became clear OfCom had been buying equipment from the new businesses modifing it and putting it into the Pirate World in new Ghosts.

    It turns our the idea was to discredit and finally prosecute thus destroy the quite legal and legitimate companies providing not just employment but education to people in the UK.

    So Clive Corrie went back to his old tricks,

    https://www.standard.co.uk/news/london/four-hundred-pirate-radio-stations-shut-down-in-london-in-just-two-years-10362974.html

    Unfortunately for Clive Corrie I can prove he purjured himself in court as did others, manufactured and provided false evidence, tampered with evidence, and intimidated witnesses.

    And I’m not the only one who can do this so it’s a stand off situation for OfCom they know there is an avalanche of realy bad news waiting to drop, as do I suspect the Met Police.

    [1] You can find multiple refrences to Clive Corrie of OfCom on the Internet where others have investigated him and found he was happy to lie and perjur himself, extort others all for “Political Purposes”. He has a linked in account where he describes himself as a consultant. My advice is best avoid, his past record is nowhere near as shiny as he portrayes.

    JonKnowsNothing January 12, 2024 11:46 AM

    @Bob, All

    re: Surviving v Living

    Humans have different versions of what each of these words means and the definitions vary by era and circumstance.

    Animals however, know exactly what it means. They have no interest in Java at *$.

    Ants know what it means. Watch an ant as you try to squish it. It runs for its life the best way it can.

    The discussion was about the possibility of evading a mass cull. These events happen globally, all the time. (1) The location varies. The outcomes are generally fatal. Few survive. Survivors have harrowing tales of how they evaded the cull. The non-survivors are not worried about Java at *$.

    Massed groups are disliked by LEOs. One tactic they use to stop grouping is called “Kettling”. (2) It corrals up everyone in the designated area. It’s a mass round up. In some countries those rounded up disappear. In others, they get prison sentences and/or fines.

    If you want to Live, you first have to Survive.

    • The odds are never in your favor

    ===
    1)
    https://en.wikipedia.org/wiki/Kettling#By_country

    Australia
    Brazil
    Canada
    Denmark
    Finland
    France
    Germany
    Israel
    Spain
    United Kingdom
    Miners’ strike, 1984
    Parliament Square disability rights demonstration, 1995
    N30 anti-WTO demonstration, 1999
    May Day 2001
    G8 summit, 2005
    G20, 2009
    Student protests, 2010
    Anti-Cuts protests, 2011
    Legal challenges
    United States
    Anti-Globalization movement, Washington D.C., 2002
    Iraq War protest, Chicago, 2003
    Republican National Convention, New York City, 2004
    Occupy Wall Street, 2011
    Presidential inauguration of Donald Trump, Washington D.C., 2017
    St. Louis, 2017
    George Floyd and Breonna Taylor protests, 2020
    Brooklyn Center, Minnesota: Daunte Wright protests, 2021

    2)
    h ttps://en .wikipedia.org/wiki/Kettling

    • Kettling (also known as containment or corralling) is a police tactic for controlling large crowds during demonstrations or protests. It involves the formation of large cordons of police officers who then move to contain a crowd within a limited area. Protesters either leave through an exit controlled by the police, leave through an uncontrolled gap in the cordons, or are contained, prevented from leaving
    • The tactic has proved controversial, in part because it has resulted in the detention of ordinary bystanders

    Bob January 12, 2024 12:09 PM

    @JonKnowsNothing

    I’ve been in crowds that have kettled, bro. It only works on crowds that aren’t fighting for their lives.

    Clive Robinson January 12, 2024 12:27 PM

    @ Winter, ALL,

    “This lie also diverts the attention from all the Republican politicians and businesses, not least the farmers, who profited from illegal immigrants. Because, these immigrants are coming because American voters, not least Republican voters in the South, employ them in masses.”

    Take the word “Republican” out and it would be as true.

    Whilst Republicans certainly do lie and raise all sorts of horror stories about “illegal immigrants” the simple fact is the “illegal immigrants” are needed to do a whole variety of jobs, that good old Americans won’t any longer do.

    The thing is human labour especially child labour is cheap especially when needed only a couple of weeks a year. They will not be replaced by machines or robots any time soon if ever.

    One of the first things I started teaching my offspring was when shopping how by look, feel and smell to tell what all raw food was the best to buy.

    Especially in these times where meat packers don’t cut off excess fat off of joints, they actually pack extra fat in –they’ve cut off of areas of the carcus destined to be chops etc– before vacuum packing a joint to significantlt increase the profit. It’s the same as in bacon, ham, chicken and turkey where 50% of the weight is water and in many cases a protein slury made from animal skins to hold the water in much like wallpaper paste.

    The thing is fat goes hard a lot quicker than muscle does thus picking a joint from the chiller and squeasing and rolling it in your hand you can tell those joints that have way to much fat in.

    Unlike a relitive of mine who puts her thumb nail through the plastic of such a pack when putting it back in the chiller to reduce the chance it will be sold at full price thus take the profit out of the game. I just put it back.

    Simmilar tricks are done with all raw foods where they can get away with it.

    But the point is our “machine sensors” are not good enough in most ways to replace a human hand, eyes or nose. And those that are sensitive enough are too large, fragile, expensive to be used or do excessive damage to the food stuff.

    As for grippers to remove delicate fruit from vines, stalks, and branches, the human hand is still way way better, and less extensive and runs on little or no energy in comparison to a mechanical system.

    The only places humans get replaced is with the likes of grains of uniform hight etc where the likes of a combined-harvester does reap, thresh, bail on mass by an extrodinary simple series of “force multipliers”.

    Which is why the likes of the hydroponics and “Genetically Modified Organisms”(GMO) get more practical investment than to replace human pickers.

    One of the scares that amuses me is “AI is after our jobs”. Whilst it is true it’s not the labouring jobs that “Illegal Immigrants” fill that Ammericans won’t do that are threatened. Or for that matter the near artisanal skills of craftsmen and similar. It’s actually the middle to upper middle class white collar semi to fully educated traditionally high paid knowledge based workers.

    As they actually form the tax base of the US and many first world nations these days of “Out Sourcing” and “Off Shoring” that Neo-Con mantra and GOP types have been pushing for the past fourty years or more…

    It’s not hard for outsiders to see where those turkeys are going to go home and roost…

    As @JonKnowsNothing keeps pointing out “the poor laws” are very much not those of half a century or more ago. The modern “poor laws” are aimed at those who have assets but not jobs to fund defending them.

    One of the things MAGA does is it’s aimed at turning the physical labouring classes into vilifing the knowledge workers so that getting rid of knowledge workers becomes almost a “public duty” (have a look at Pol-Pot and friends with the Cambodian genocide “killing fields”). And much like the burning of the books etc allowing brain dead mostly cowardly fascists and the corrupt to appear as “Heroic Strong Men”, and we only have to look at Putin and Co to know where that goes in more contemporary times.

    Bob January 12, 2024 12:45 PM

    @Clive

    A lot of American Conservatives actually like Putin. Back when Obama was president, one of our loudest, most prolific conservative pundits was all but masturbating on air over the “shirtless Putin on a horse” pictures that were making rounds. For contrast, he railed about Obama wearing a helmet while riding a bicycle.

    This was years before MAGA.

    Clive Robinson January 12, 2024 4:55 PM

    @ Bob,

    Yes I remember both Putin’s “Homo-erotic Machismo” nonsense, as in UK cultural parlance it made him look like a “Chinless Willy-Woofter”… But then “Brokeback Mountain” came out of US culture and made it big in 2006, with one or three snide political jokes thrown in at the time.

    So you had Conservatives ranting into hyperventilation over Brokeback yet as you say drooling over Putin…

    Where I come from that’s called “mixed messages” and implies “closet behaviours” but then we are talking about those famed “Dickpic Senators and their aids” who feel that they are above it all, self entitled etc etc.

    And if they did not have those peccadillos of monstrous and career ending proportions, which sponsor would trust them to do as they were told…

    US Conservative politics, “No well stuffed closet, no career”.

    I think you can see why outside observers look confused and shake their heads in disbelief.

    Bob January 13, 2024 1:01 AM

    @Clive

    A lot of us are shaking our heads as well. People who live in rural states still get more voting power in the US than do people in higher-density areas. It’s codified into our Constitution at the federal level, which has actually proven quite difficult to amend absent some pretty horrific bloodshed and/or religious fervor.

    JonKnowsNothing January 13, 2024 3:18 AM

    @Bob, @Clive, All

    re: rural states still get more voting power

    (USA) This is not exactly accurate and in might mislead others about how the US Government, Federal and State, is organized.

    Regardless of party issues, our Congressional Branch is divided into 2 sections: The House of Representatives and The Senate. (1)

    • Representatives are elected by population distribution. So rural states do have fewer Representatives than populous states. There are 435 Representatives. (2)
    • Senators are apportioned 2 Senator for each state. So rural states have just as many Senators as populous state. There are 100 Senators.

    Our system is somewhat based on the UK system but we do not have a Parliament.

    Our modern Senate is elected, which was not always the case. (3) The functions of the Senate have undergone a number of rule changes and alterations. Unlike the House of Lords, the US Senate has an active and important legislative function that is required for the other 2 branches of government, US Executive (President) and Judiciary (SCOTUS).

    States have their own versions and some have different names, titles or styles but the basic organization is similar to the Federal Government setup.

    County, City organizations mirror their state layouts.

    Each state has their own judiciary which works differently than the federal courts, as each state enacts their own laws (ex Federal ROE laws, State ROE laws).

    ===
    1)

    ht tps:/ /en.wikipedia.org/wiki/United_States_Congress

    2)
    htt p s:// en.wikipedia.org/wiki/Non-voting_members_of_the_United_States_House_of_Representatives

    There are some non-voting members due to US views of our existing colonial territories and the status of Washington District of Columbia the US Capitol city.

    • There are currently six non-voting members: a delegate representing the District of Columbia, a resident commissioner representing Puerto Rico, as well as one delegate for each of the other four permanently inhabited U.S. territories: American Samoa, Guam, the Northern Mariana Islands and the U.S. Virgin Islands. A seventh delegate, representing the Cherokee Nation, has been formally proposed but not yet seated, while an eighth, representing the Choctaw Nation, is named in a treaty but has neither been proposed nor seated.

    3)
    ht tps:/ /en .wikipedia.org/wiki/History_of_the_United_States_Senate

    Historical changes to the make up and function of the US Senate.

    Bob January 13, 2024 3:26 PM

    @JonKnowsNothing

    The net effect of the Senate itself, and the Electoral College, is to give outsized voting power to rural dwellers. It’s a result of our Constitution being a compromise between wildly different states, many of which included slave states.

    It is exactly accurate, and that’s because the US was founded on compromise with slavers.

    JonKnowsNothing January 13, 2024 6:01 PM

    @Bob, All

    re: outsized voting power to rural dwellers

    Only partially accurate.

    The mass power lies in the Agricultural Committees in both the House and Senate. You cannot really separate Farm Policy from anything else because contrary to what Techies like to think, Farming Rules The Roost, in Federal, State and County policies.

    City polices don’t really affect the Farmers much.

    The Electoral College has a very small role that becomes a big role every 4 years. They haven’t got much else to say in between.

    The monetary purse is held by The House, not the Senate. If you want $$$ you have to go to the House where city dwellers control the flow. If you want to ratify a treaty you go to the Senate where power is divided equally between all the states.

    Your attempts to derail the context with the issue of “slave state reparations” omits that these issues do not reside solely on one side of Congress, it is both sides of Congress: House and Senate. The many thorned issues affect both Urban and Rural Dwellers. The on-going discriminating(s) in State Policies and Federal Polices and County Polices maintain the effects of 200yrs of land appropriations and urban disenfranchis-ement. (1)

    This aspect of City v Urban conflict of interest is not just an issue in USA, nearly every country has such conflicts. The results of these conflicts are long lasting both in human terms but also in ecological terms.

    • It ain’t over until it’s over, and it ain’t never over

    ===

    1)
    HAIL Warning

    ht tps:/ /ww w.theguardian.com/us-news/2024/jan/13/mississippi-child-school-food-program-welfare-state

    • Mississippi quits child food program amid Republican ‘welfare state’ attack
    • Mississippi is withdrawing from a federal program to feed children during their summer break from school that would give electronic benefit transfer (EBT) cards to low-income families to supplement food costs when academic classes are out of session [$40 USD/ month for 3 months]
    • [Mississippi is] the poorest state in America choosing to turn down federal aid for poor children to eat
    • 12 States turned down the summer food program:
      • Alabama, Oklahoma, Alaska, Florida, Mississippi, South Carolina, South Dakota, Georgia, Louisiana, Texas, Vermont and Wyoming.

    JT January 15, 2024 11:45 AM

    I know there’s a wise guy in every group, but I distinctly dislike the term, democracy. There is a reason the Founding Fathers never used the word in any major document. A democracy is majority rule, period. There are no minority rights, like the Bill of Rights. We are supposed to be a constitutional republic, even if the Constitution is ignored and violated every day inside the Beltway. Sadly, I doubt most college age students have ever read the Constitution.

    As for improving our current form of government, we need more major parties. The two party system leaves many of us cold. Many of us view them as two branches of the same uniparty. Many of us don’t even bother to vote, because it really doesn’t matter. We will have more war, more spending, and more debt, regardless.

    Weby corp February 2, 2024 7:46 AM

    Agree with what Bownse said, if literally EVERYONE could vote on any issue – we’d be doomed really quick. Electing representatives that are actually savvy in those questions is the way to go.

    Chris Drake March 16, 2024 2:44 AM

    “AI” is nothing more than a search engine with amazing context – there’s absolutely nothing “intelligent” whatsoever about it, and it cannot “reason” or in any way behave like a real person: the only thing it can do is reconstruct sentences that fool stupid (and many smart) people into overlooking the fact that’s it’s merely regurgitating tokens from a training set.

    Someone also chose that training set (typically censoring difficult issues), and someone fine-tuned and “aligned” it (always enforcing thorny-issue avoidance rules, and who knows what else) before release.

    You can’t use or trust anything it outputs, least of all in any governance roles!

    Leave a comment

    Login

    Allowed HTML <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre> Markdown Extra syntax via https://michelf.ca/projects/php-markdown/extra/

    Sidebar photo of Bruce Schneier by Joe MacInnis.