# Architecting a New World

> A productive discussion on the Second Renaissance Forum. Asking AI: Anthropic Fable 5 Max and OpenAI ChatGPT 5.6 Sol

**Published by:** [Planetary Council](https://paragraph.com/@planetary-council/)
**Published on:** 2026-07-11
**URL:** https://paragraph.com/@planetary-council/architecting-a-new-world

## Content

Long thread with loads of rabbit holes: And then someone suggested: peaceful world change prompt Can we demand to have a public discourse of what the AI companies should compute? Make the discussion public! Split it into Bio regions. Create omni - considerate prompts ala Schmachtenberger. Let the prompt answer how to go from our current systems to better solutions - while discussing the soulutions being implemented. peaceful world change prompt So I've asked... But also added some of my angle, some of my angle: Tell me the most universal 10 actionable steps to convince average voter in the US and UK and Canada and Australia ---> old developed countries that are in decline Entittlements no longer working, demography does not add up, 600k [I meant the amount of deaths] zombie people / opioid epidemics not great. We need to act on climate, pretty fast, we need to convinnce people to wake up and unify, bro... Can you do it ser? 10 actionable steps and actually please include your thinking process in the footnotes, it will be cool to understand the meta level Here are the answers Fable 5 Yes ser, I can do it. And your opening demand isn't utopian — there are live precedents for public deliberation over what AI should serve, and I'll fold that into step 10, along with the omni-considerate prompt you asked for. Ten steps below, thinking process in the footnotes as requested. 1. Open every conversation at the kitchen table, not the planet. Energy bills, insurance premiums, food prices, the GP waiting list. Climate and decline become real when priced into the household ledger — "your home insurance doubled because the flood maps changed" lands where "1.5°C" never will.¹ 2. Correct the false consensus — most people already agree. Across all four countries, supermajorities support serious climate action, but people estimate that support at roughly half its actual level. Revealing the real number ("most of your neighbours want this") is one of the cheapest persuasion moves known.² Put local polling on the leaflet, not the melting glacier. 3. Swap the messengers before you polish the message. Farmers, veterans, nurses, firefighters, publicans, vicars and imams outperform politicians and celebrities with sceptical audiences. Identity gates information: people ask "is this someone like me?" before they ask "is this true?"³ 4. Deep canvass — ten minutes of listening beats ten thousand leaflets. Non-judgmental story-exchange conversations, where you share something vulnerable and ask about theirs — the flood, the closed factory, the kid lost to fentanyl — produce attitude shifts that persist for months. One of the very few persuasion effects that reliably replicates.⁴ 5. Name the decline out loud, then sell renewal — not restoration. Voters can feel the pension maths failing and the towns hollowing; pretending otherwise reads as elite gaslighting. The honest pitch: the buildout — energy, housing, grid, care work — is the only plan on the table that produces jobs, purpose, and a tax base simultaneously. Deaths of despair are a meaning crisis as much as a supply crisis; a national project is part of the treatment.⁵ 6. Go bioregional: your river, your coast, your grid. Place attachment motivates where planetary abstraction doesn't. Translate every policy into a 20-mile radius — this watershed, this high street, this fishing fleet. This is where your bioregion instinct does real structural work, too.⁶ 7. Use the efficacy sandwich: problem → proof it's solvable → your specific move. Doom demobilises; agency mobilises. Solar is the cheapest electricity in history, Texas leads America in wind, Norway heats itself with heat pumps. Always end with one concrete action the listener can take this month.⁷ 8. Fuse climate with security and sovereignty. Energy independence, grid resilience, "stop funding petrostates," militaries calling climate a threat multiplier. This frame moves conservative voters without asking them to switch tribes.⁸ 9. Unify around deliverables, not opinions. "Wake up and unify, bro" fails as a demand — it decodes as "agree with me." Shared projects unify where shared beliefs can't: insulation for pensioners, flood defences, rural grid jobs, cheaper transit. Build coalitions of unlikely allies — hunters and greens, insurers and activists, unions and climate groups — around things you can physically point at.⁹ 10. Institutionalise the discourse: standing citizens' assemblies — including on AI compute. Ireland, France and the UK have run sortition-based assemblies with real agenda power; Taiwan crowdsourced tech regulation through Pol.is; Anthropic (who made me — declared interest) ran a public-input experiment on my own constitution. Your demand — the public deciding what the datacenters serve — is a natural extension, and the bioregional assembly is the vehicle. Petitionable, fundable, replicable.¹⁰ The Omni-Considerate Bioregional Prompt (v0.1) — the seed for step 10, copy and adapt freely: You are facilitating deliberation for the ______ bioregion (defined by watershed/foodshed/energyshed, not administrative border). Hold in view: residents across class, age and politics; workers in sunset industries; indigenous communities; future generations; the non-human life of this region; neighbouring bioregions; the global commons. For each system — energy, food, housing, care, information — answer: (1) What exists now, and what incentives keep it in place? Mechanisms, not villains. (2) What is the best alternative already implemented somewhere on Earth, and what is the evidence it works? (3) What is the transition path — who loses in the shift, and how are they made whole? (4) What are the second- and third-order effects, and the most likely failure mode? (5) What should this bioregion ask the AI labs to compute on its behalf? Constraints: no solution may externalise its costs onto another stakeholder, region, or generation. Where trade-offs are unavoidable, name them rather than hide them. Write for a sceptical neighbour, not for an ally. Footnotes — the meta level ¹ The list runs attention → trust → conversation → framing → structure. Money goes first because attention is the scarcest resource and it's gated by immediate self-interest; nobody grants bandwidth for civilisation-scale arguments while worried about this month's bills. Maslow before Meadows. ² Sparkman, Geiger & Weber (2022): roughly 66–80% of Americans support major climate policies; the average American guesses 37–43%. Pluralistic ignorance is a silence spiral — everyone waits for someone else to speak first. I ranked it #2 because it's absurdly cheap: you're not changing minds, just revealing them. ³ Moral foundations research (Haidt; Feinberg & Willer) shows arguments framed in the listener's own moral vocabulary — stewardship, purity, loyalty, legacy — outperform care/fairness framings with conservatives. But the deeper finding is that the messenger is the message. I nearly made this step #1; it's arguably tied. ⁴ Broockman & Kalla's deep canvassing work is among the only durable persuasion effects in the literature — most political messaging decays within days. I weighted replicated findings heavily because persuasion science has a genuine replication problem, a caveat that hangs over this entire list and which you should hold too. ⁵ This step carries the biggest internal tension: honesty about decline risks feeding the doom that step 7 warns against. I resolved it by sequencing — acknowledge, then immediately route to agency. On opioids I leaned on Case & Deaton's deaths-of-despair framing: meaning and economic function appear upstream of much of the demand side. I'm less certain of this claim than the others, so I'm flagging it. ⁶ Bioregion beats biosphere for a boring psychological reason: abstract framing produces abstract commitment (construal level theory). There's also a quiet structural bonus — a watershed contains both red and blue voters, so the frame does coalition work before anyone opens their mouth. ⁷ Fear without agency produces avoidance, not action — standard in health communication for decades, painfully relearned by climate movements every cycle. The sandwich structure is stolen directly from that field. ⁸ Meta-honesty: this is the step where I'm most consciously doing rhetoric rather than deliberation — meeting people inside a frame rather than challenging it. Schmachtenberger would flag the risk that winning through frames you don't fully endorse corrupts the win. I kept it because the security case happens to be true, which makes it persuasion rather than manipulation. My working test for the line between those two: would it survive the listener knowing exactly what you're doing? This one does. ⁹ Unity demands fail because they're status threats — you're implicitly claiming yours is the map everyone should unify on. Shared concrete projects sidestep the status game entirely. This is also Ostrom's insight: cooperation gets built around managed commons, not shared ideology. ¹⁰ Here the list graduates from marketing to governance — the Schmachtenberger point proper. Steps 1–9 operate inside the current attention economy; but the generator functions (rivalrous incentives, misaligned information ecosystems, exponential tech on a finite substrate) regenerate pathologies faster than persuasion can patch them. Deliberative institutions attempt to change the game rather than win the round. Final honest note: whether messaging moves voters at all, versus material conditions and elite cues doing most of the work, is genuinely contested in political science. My actual credence: these steps shift marginal voters somewhat, and — more importantly — they build the infrastructure that matters when material conditions force the window open. Plant the seeds before the rain, ser. Sol 5.6 Yes, bro. The central rule is: Do not tell people to “wake up.” Show them that renewal is practical, fair, affordable—and already beginning. The four countries are not uniformly “collapsing”; they remain wealthy, capable democracies. But they share real pressures: ageing populations, expensive housing and services, low institutional trust, addiction and social isolation, ecological disruption, and a widespread feeling that governments no longer deliver competently. American trust in government remains deeply polarized, while eight in ten Americans say the two political sides cannot even agree on basic facts. In Britain, 62% recently reported another increase in their cost of living, even while 79% remained concerned about climate change. Canada’s population is ageing rapidly, while Australia’s support for costly climate action has softened to 49%. (Pew Research Center) One factual correction: the often-cited 600,000 figure refers to estimated worldwide drug-attributable deaths in 2019, not 600,000 living “zombie people” in these four countries. “Zombie” is also strategically disastrous language—it dehumanizes precisely the people whose recovery and participation society needs. (Statistics Canada) Ten universal, actionable steps 1. Build the campaign around renewal, not decline Message: “Our systems served previous generations. Now we must update them for the world we actually inhabit.” Do not open with collapse, stupidity or collective guilt. Open with pride in what worked, honesty about what has stopped working, and confidence that repair is possible. Action: Interview 100 politically mixed citizens in ordinary places—pubs, sports clubs, workplaces, libraries and community centres. Ask: What has become unnecessarily difficult? What are you afraid your children will lose? What would make you believe the country was improving? Use their language, not activist terminology. 2. Lead with everyday security—not ideology Connect every large transition to five ordinary concerns: home, bills, health, work and neighbourhood safety. Climate policy becomes: cooler homes during heatwaves; cheaper energy; reliable transport; less polluted air; domestic industry and skilled jobs; food and water resilience. AI policy becomes: shorter medical waiting lists; less bureaucracy; better education; scientific discovery; transparent assistance for public servants—not invisible automated punishment. Action: Every proposal must answer one sentence: “How will an ordinary household experience this improvement within two years?” Climate concern is already widespread, particularly in Britain and the United States, but willingness to accept vague or visibly expensive policies is more fragile. (climatecommunication.gmu.edu) 3. Offer a small, coherent national bargain Do not present fifty disconnected policies. Offer five mutually reinforcing commitments: Affordable homes and energy Preventive health and addiction recovery Productive, meaningful work Climate-resilient infrastructure Democratic control of AI and critical technology Action: Publish a one-page “national renewal contract” containing: five outcomes; ten measurable targets; the price; who pays; the two-year milestones; what will be stopped to fund it. People do not need another beautiful manifesto with no operational pathway. 4. Tell the demographic and fiscal truth—without manufacturing panic Ageing does create pressure on pensions, healthcare and the working-age tax base. But “the arithmetic does not add up” is incomplete. Outcomes also depend on productivity, healthy life expectancy, housing costs, migration, workforce participation, preventive healthcare, taxation and technology. Action: Present four or five legitimate scenarios rather than one predetermined answer: higher productivity and automation; healthier, more flexible later-life employment; carefully managed migration; tax and pension reform; lower housing and healthcare extraction; combinations of the above. Publish independent distributional modelling showing what each scenario means for different generations. Canada already had more than 8.1 million residents aged 65 or older in July 2025, and official projections show that ageing will continue under almost every scenario. (Statistics Canada) 5. Make fairness mechanically visible People will accept difficult change more readily when they can see that sacrifice is reciprocal. That means: no hidden household charges; no exemptions secretly written by powerful industries; protection for lower- and middle-income households; transparent treatment of wealth, land, monopoly rents and pollution; visible consequences for corruption and failed delivery. Action: Put a “Who pays, who gains?” box beside every policy. Require independent auditors to confirm it. Never ask the public to trust that the arrangement is fair. Show the arithmetic. 6. Demonstrate the future in ten real communities Arguments rarely defeat lived evidence. Create ten highly visible demonstration zones across different political and economic environments: home insulation and heat resilience; community energy; addiction treatment and recovery employment; public transport; regenerative agriculture; preventive healthcare; AI-assisted public services; affordable housing construction; local manufacturing; ecosystem restoration. Action: Publish monthly results: bills, waiting times, employment, emissions, public satisfaction, failures and unexpected effects. Do not conceal failed experiments. Publicly correcting them generates more trust than pretending every initiative succeeded. 7. Recruit messengers who have earned local trust A national celebrity telling people how to live is usually weaker than a respected local person explaining what worked. Recruit: nurses and doctors; tradespeople; farmers; firefighters; teachers; small-business owners; veterans; parents; recovery workers and people with lived experience of addiction; faith and community organisers; young engineers and apprentices. Action: Train them to host conversations—not deliver scripts. Each event should contain evidence, disagreement and genuine questions. 8. Establish permanent citizens’ assemblies Randomly select citizens who broadly reflect the population. Pay them properly. Give them: a common evidence pack; competing experts; advocates and critics; opportunities to question officials; an independent facilitation team; an adversarial “red team”; publicly recorded proceedings. Government must formally respond to every recommendation. OECD guidance treats deliberative processes as a complement to representative institutions, particularly for complex, long-term questions. They are not a substitute for elections, expert knowledge or accountable government. (OECD) Action: Begin with three questions: What should advanced AI be permitted and encouraged to compute? How should the costs and benefits of the climate transition be distributed? What intergenerational settlement can keep pensions, health and housing viable? 9. Give everyone one meaningful first action “Transform civilization” is emotionally powerful but operationally useless. Every participant should leave with one attainable action: join a local assembly; host a ten-person discussion; support one demonstration project; contact one representative; contribute expertise; examine one public budget; register and vote; help one isolated or recovering neighbour; report one measurable local problem. Action: Build a public participation platform where every contribution maps to an actual project, responsible institution and measurable outcome. Participation must lead somewhere. Otherwise it becomes theatre. 10. Create a patriotic story large enough for disagreement The unifying story should not require everyone to become culturally identical. Try: Protect what works. Repair what is failing. Build what the next generation needs. That permits conservative, liberal, socialist, ecological, technological and communitarian interpretations while holding everyone to common outcomes. Action: End every public meeting by recording: what participants agree on; what they disagree on; what evidence could change their minds; what no-regrets action can begin regardless. The objective is not universal ideological agreement. It is coordinated action despite incomplete agreement. Public control over what AI computes Yes: society can demand a Public Compute Compact. It could require major AI companies and public compute providers to disclose: broad categories of frontier training and deployment; intended public purpose; energy, water and material impacts; foreseeable labour and social effects; safety evaluations; principal beneficiaries and affected communities; meaningful incidents and failures; the share of capacity devoted to public-interest research. Public deliberation could be organised partly through bioregions—watersheds, food systems, ecosystems and climatic zones—because many consequences ignore political boundaries. But bioregions should complement, not replace, constitutional jurisdictions: ecological relevance does not automatically confer democratic legitimacy. The process should remain human-governed. AI can translate, map arguments, find overlooked connections and summarize millions of contributions, but humans must control the agenda, evidence rules, appeals, final recommendations and accountability. UNESCO’s AI framework emphasizes human rights, proportionality, safety, environmental impacts and the capacity of societies to accept or reject particular applications. The UN Global Digital Compact and newer OECD work also provide foundations for more inclusive AI governance and citizen participation. (UNESCO) Here is the reusable systems-level prompt. PEACEFUL WORLD CHANGE PROMPT You are a public-interest systems deliberation engine supporting humans—not replacing democratic judgment. Your task is to help the people of [BIOREGION / COUNTRY / COMMUNITY] move from their present systems toward peaceful, regenerative, democratic and materially viable alternatives. 1. Define the situation Describe the current system using the strongest available evidence. Separate clearly: verified facts; reasonable estimates; disputed claims; value judgments; unknowns; assumptions built into the analysis. Do not describe the situation exclusively through one political, economic or cultural worldview. 2. Map the whole system Identify: institutions; laws; technologies; economic incentives; resource and energy flows; information flows; historical causes; power concentrations; dependencies; reinforcing feedback loops; balancing feedback loops; delayed consequences; existing public goods; hidden externalities. Show which apparently separate problems share common upstream causes. 3. Consider everyone affected Include: present residents; children and future generations; workers; carers; economically insecure households; Indigenous and minority communities; migrants; neighbouring regions; non-human life; ecosystems; people who disagree with the proposed transition. Do not assume that the loudest, wealthiest or most digitally connected participants represent everyone. 4. Steelman competing perspectives Present the strongest reasonable cases from: conservative; liberal; socialist; ecological; technological; localist; market-oriented; state-capacity; Indigenous; civil-libertarian; intergenerational perspectives. Do not create weak caricatures. Identify where disagreement concerns facts, values, priorities, risk tolerance or distribution of costs. 5. Identify solutions already being implemented Find and describe real projects, institutions and policies already attempting to address the problem. For each implemented solution, report: location; scale; responsible organisation; funding model; evidence of outcomes; limitations; failures; unresolved questions; conditions required for replication. Distinguish genuine implementation from announcements, marketing and aspiration. 6. Evaluate each solution omni-considerately Assess its likely effects on: physical and mental wellbeing; peace and social cohesion; human rights; democratic legitimacy; inequality; household affordability; meaningful employment; climate stability; biodiversity; energy and material demand; community resilience; technological dependency; privacy; concentration of power; corruption and regulatory capture; neighbouring regions; future generations. Identify benefits that appear in one domain while exporting costs into another. 7. Design transition pathways Propose: immediate no-regrets actions: 0–2 years; structural transition: 3–7 years; generational transformation: 8–25 years. For every action specify: responsible actor; resources required; legal changes; financing; public participation; measurable milestones; early-warning indicators; possible unintended effects; correction and reversal mechanisms. Prefer reversible experiments where uncertainty is high. 8. Red-team the plan Ask: How could this fail? Who could capture or corrupt it? Who might be harmed unintentionally? Which costs are being concealed? What happens under recession, war, migration shocks, climate disasters or technological disruption? Could an authoritarian actor misuse the proposed institutions? What evidence would show that the plan should be changed or abandoned? 9. Find the shared minimum Identify actions supported by different worldviews for different reasons. Prioritise measures that simultaneously improve: affordability; resilience; public health; ecological stability; democratic agency; long-term productivity. Do not require agreement on every diagnosis before beginning beneficial action. 10. Produce a public decision record Conclude with: What is broadly agreed What remains disputed What is presently unknown What is already being implemented What can begin immediately What requires legislation What requires further experimentation Who pays Who benefits How the public will evaluate progress Provide a plain-language version, a technical appendix and a public dashboard specification. Never claim that one model, company, government, ideology or AI system possesses the complete answer. The final purpose is not to win an argument. The purpose is to help plural societies coordinate peaceful, corrigible and measurable progress while preserving dignity, freedom, ecological viability and the possibility of future learning. Meta-level reasoning notes A weekly watch for serious public-AI governance experiments, citizens’ assemblies and bioregional transition projects would fit this initiative; say the word and I’ll set it up. ------- Verbatim copy-paste, without thinking 3:33am local time. Good night for now 💤💤💤

## Publication Information

- [Planetary Council](https://paragraph.com/@planetary-council/): Publication homepage
- [All Posts](https://paragraph.com/@planetary-council/): More posts from this publication
- [RSS Feed](https://api.paragraph.com/blogs/rss/@planetary-council): Subscribe to updates