Redistributed

General

Doc Review: AI and Human Civilization — Strategic Foresight

Published March 19, 2026

doc-reviewforesightgovernanceinequalitylaborai-safety

Document: "AI and Human Civilization: Strategic Foresight" — Open letter by Renaud Bougueng T. Source: Maisons Noaré Published: March 18, 2026

Every few decades, someone tries to connect all the threads — technology, economics, governance, human nature — into a single coherent picture of where civilization is heading. Yuval Noah Harari did it for the past in Sapiens. Renaud Bougueng T., CEO of Maisons Noaré and Rauio, attempts it for the future in this open letter on AI as civilizational transformation.

The letter is ambitious, genuinely wrestling with species-scale questions. It synthesizes strategic foresight, innovation economics, and speculative governance into a single sweeping argument. Not everything lands equally, but the attempt matters — because the ideas it raises deserve to be connected, extended, and built upon. That's what this review tries to do.

The Core Argument

Bougueng's thesis runs through four interlocking claims, each reaching toward something important:

1. Technological convergence is outpacing institutions. AI, quantum computing, biotech, robotics, and gene editing are converging faster than political, social, and cultural systems can adapt. He calls this gap "institutional decoherence" — a term that captures something real. It's not just that regulation is slow; it's that political, cultural, economic, and social institutions are adapting at different rates, creating gaps that powerful actors exploit. The EU AI Act took years to pass and is already struggling with implementation (The EU AI Act Takes Effect: What's Actually Changing). US executive orders on AI proved reversible overnight (AI Executive Orders: What Changed and What Didn't). International AI summits produce communiqués, not binding governance. Bougueng names the pattern precisely.

2. We're entering a "two-species civilization." Bougueng treats AI as an ontological species — computationally cognitive machines coexisting with biological humans. This is the letter's most provocative framing, and it opens up genuinely interesting governance questions: if AI systems increasingly act with autonomy, what frameworks do we need for human-to-AI and AI-to-AI relations? He invokes Geoffrey Hinton's suggestion that superintelligent systems may need "something akin to a maternal instinct," then reverses the frame — humans are already parenting AI on our collective memory.

3. Material production will commoditize. Full automation of safety, shelter, food, and industrial goods could free most people from material labor, opening space for what Bougueng frames in Maslovian terms: love, belonging, esteem, self-actualization. UBI serves as a "temporary bridge" to automated abundance. New "post-economies" emerge around authentic human experience, potentially powered by cryptographic value systems.

4. Intelligence is becoming a class divider. Despite projected abundance, cognitive capacity won't be equally distributed. "Purchasing power now buys cognitive escape velocity" — frontier AI agents, dedicated compute, continuous augmentation compound advantage exponentially. Lower-income groups get basic-tier models and active discrimination in opportunity.

What Makes This Letter Worth Engaging With

A Map, Not Just a Warning

Most writing about AI's civilizational impact falls into two camps: breathless optimism or existential alarm. Bougueng tries something harder — drawing a map that shows both the dangers and the openings. Like Sapiens, the letter asks readers to zoom out far enough to see patterns that are invisible at ground level. The institutional decoherence concept, for instance, isn't just a complaint about slow regulation. It's a diagnostic that helps explain why governance keeps failing, across multiple domains simultaneously.

This kind of big-picture thinking has practical value. Policymakers need frameworks that connect AI governance to broader institutional questions. Technologists need context for the systems they're building. Citizens need stories that make sense of change without either flattering or terrifying them.

The Inequality Mechanism Is Precisely Named

"Purchasing power now buys cognitive escape velocity." This may be the letter's single sharpest observation. It captures a dynamic that Korinek's research on AI's distributional effects (Paper Review: The Distributional Effects of AI) has documented in economic terms — AI productivity gains flow disproportionately to capital owners and high-skill workers — but gives it a visceral, immediately graspable name. High-income individuals and entities using frontier AI agents with dedicated compute don't just have better tools; they accumulate advantage at exponential rates.

Bougueng extends this to a bio-computational spectrum: within two decades, cyborgs and hybrid humans may create entirely new class categories. Whether or not you accept the timeline, the underlying logic — that differential access to cognitive augmentation will compound existing inequality — deserves serious policy attention now, before the dynamics become entrenched.

Three Modes of Agency

The letter identifies three agency dynamics: humans directing AI, AI directing humans, and AI directing other AI. This taxonomy is useful because it maps real territory. Algorithmic management already directs millions of workers — ride-share drivers, warehouse workers, content moderators all take instructions from systems (The Manager Is an Algorithm). The third mode — AI-to-AI coordination — is the one Bougueng flags as approaching "escape velocity," where humans become the operational bottleneck. Agentic AI systems are already moving in this direction (Agentic AI and the Autonomy Question), and having clear language for these modes helps us think about governance at each level.

Where the Ideas Need More Architecture

The letter's ambition is also its vulnerability. Several claims arrive as bold assertions that deserve to be developed into full arguments with supporting structure. This isn't so much what the letter gets wrong as where it opens doors it doesn't yet walk through.

The Species Metaphor: Provocative, Needs Grounding

The "two-species" framing is the kind of idea that starts productive conversations — but it also carries risks if taken too literally. AI systems don't yet have interests, survival instincts, or agency independent of the humans and corporations that build and deploy them. The metaphor could inadvertently naturalize what are actually political and economic choices — as Kate Crawford documents in Atlas of AI, AI is not an autonomous force but an extractive industry with material supply chains and concentrated ownership.

The more interesting version of this idea — and the one worth developing — is that we need governance frameworks for entities with functional autonomy, even if they lack consciousness. Legal systems have done this before: corporations are treated as persons for many legal purposes without anyone claiming they're sentient. What would a similar framework look like for AI agents? Bougueng gestures toward this but could push further. The governance innovation might matter more than the ontological claim.

From UBI Bridge to Institutional Imagination

Bougueng treats UBI as a "temporary bridge to automated abundance." The idea has real merit as a starting point, but it carries more weight here than it can bear alone. UBI faces genuine political difficulty, and pilot programs show mixed results. More fundamentally, Acemoglu and Johnson argue in Power and Progress (Book Review: Power and Progress) that technology never automatically benefits everyone — abundance doesn't distribute itself.

But here's what's interesting: the letter's underlying intuition — that we need transition mechanisms for a post-scarcity economy — is sound. The question is what else goes on that bridge alongside UBI. Public AI infrastructure, as Mariana Mazzucato's work on the entrepreneurial state suggests, could return public value from publicly funded research. Commons-based governance models, drawing on Elinor Ostrom's framework, could manage shared AI resources without defaulting to either corporate control or state bureaucracy. Cooperative ownership models could give workers and communities stakes in automated systems (AI Cooperatives: Alternatives to Corporate Models). Bougueng opens the door to this conversation. Walking through it means imagining multiple transition mechanisms, not just one.

Creative Destruction Deserves Its Complexity

The letter opens with the 2025 Riksbank Prize for Mokyr, Aghion, and Howitt — celebrating how "creative destruction drives sustained prosperity." The economic history here is real, but incomplete. Creative destruction creates winners and losers, and the distribution between them is a political question, not a natural law. Acemoglu and Brynjolfsson have documented this tension extensively — Acemoglu emphasizing how automation can replace workers without productivity gains, Brynjolfsson showing how deployment choices determine outcomes.

What makes this a genuinely fascinating area is that the answer isn't predetermined. Bougueng is right that technological convergence could create extraordinary abundance. The debate between Acemoglu and Brynjolfsson (The Automation Debate) isn't about whether AI creates value — it's about what kinds of AI we build and who shapes those choices. This is the space where foresight documents like Bougueng's become most valuable: they can illuminate the decision points where different futures branch.

Artificial Dissuasion: An Idea Worth Building

The letter introduces "artificial dissuasion" — a game-theoretic maturity model for preventing catastrophic autonomous AI deployments, inspired by nuclear deterrence. The concept arrives without full supporting structure, but the core question it raises — how do we create credible deterrence against dangerous AI deployment? — is one of the most important in governance today.

Nuclear deterrence worked (to the extent it did) because the logic was simple: mutual destruction. AI systems present fundamentally different dynamics — continuous deployment, dual-use applications, diffuse actors, no clear second-strike capability. This means the concept needs significant development to become operational. But rather than dismissing it, it's worth asking: what would an AI deterrence framework look like if we took the computational costs of AI chips as seriously as we take fissile material? The analogy between AI chips and nuclear materials has been explored in export control policy (AI Chip Export Controls: Technology as Geopolitics), and Bougueng's concept could connect to that work in productive ways.

Post-Labor Futures: Between Aspiration and Evidence

Bougueng envisions automation "liberating most people from mundane labor" for higher pursuits. This is the Maslovian frame applied to political economy. The aspiration is genuinely appealing — and worth taking seriously rather than simply debunking.

The complication is that automation, historically, doesn't eliminate work so much as transform and relocate it. Content moderators bear psychological trauma. Data labelers in Kenya and the Philippines earn $1-2 per hour. Freelance markets are contracting. The gap between "automation liberates" and "automation displaces" is where the most important policy work happens.

But Bougueng's vision isn't naive — it's incomplete. A fuller version would grapple with how we build the institutions that make post-labor flourishing possible. What does education look like when material labor is optional? How do communities maintain cohesion and purpose? What new forms of meaningful contribution emerge? These are questions worth designing toward, not just critiquing as unrealistic. Ursula Franklin's distinction between "holistic" and "prescriptive" technologies offers a useful lens here: the goal is AI that augments human capability and purpose, not AI that reduces human judgment to following machine outputs.

Connecting the Threads: A Map of the Territory

Step back far enough, and the ideas in Bougueng's letter connect to a larger picture of where AI sits in human civilization — one that's more nuanced than either techno-optimism or techno-pessimism:

The institutional question — Bougueng's "institutional decoherence" connects to a deeper pattern. Every major infrastructure technology (electricity, water, broadcasting, roads) went through a period where capability outpaced governance. The resolution was never automatic. It required political struggle, institutional innovation, and new frameworks for public obligation. AI is in that gap right now. History suggests the gap closes — but how it closes is the fight that matters.

The inequality question — "Cognitive escape velocity" connects to a long arc of technology-driven inequality. The printing press concentrated knowledge among the literate; the industrial revolution concentrated wealth among capital owners; the internet concentrated attention among platform companies. Each time, counter-movements emerged — public education, labor organizing, antitrust regulation — that redistributed access. The question is whether we can build those counter-institutions faster this time, before AI-driven inequality becomes self-reinforcing.

The governance question — The letter's "artificial dissuasion" concept, however underdeveloped, connects to a fundamental challenge: how do you govern a technology that moves faster than deliberation? The answer emerging from multiple fields — Ostrom's polycentric governance, Benkler's commons-based production, Mazzucato's mission-oriented policy — is that governance doesn't have to be centralized or slow. It can be distributed, adaptive, and embedded in the systems themselves. This is where the most promising work is happening.

The human question — Beneath all the policy and economics lies the question Bougueng is really asking: what do humans do when machines can do most of what we currently do for a living? This is Maslow, yes, but it's also Franklin, Mumford, and the entire humanist tradition asking whether technology serves human flourishing or replaces it. The answer depends entirely on design choices being made now.

What This Means for You

If you're in strategic foresight or policy:

  • "Institutional decoherence" is a useful diagnostic. Start using it — and start designing institutions that can adapt at technology speed.
  • The species metaphor is less useful than the functional autonomy question. Focus governance frameworks on AI agency levels rather than ontological status.

If you're concerned about inequality:

  • "Cognitive escape velocity" names the mechanism precisely. Access to frontier AI is a compounding advantage — and closing that gap is an urgent policy priority.
  • Watch for the class implications of bio-computational augmentation. This isn't science fiction — it's a 10-20 year policy question.
  • Support concrete alternatives: public AI infrastructure, cooperative models, commons-based governance.

If you're an AI developer or investor:

  • The letter's optimism about automated abundance is worth taking seriously — and worth pairing with distributional evidence. Building toward abundance and building toward equitable access are both necessary.
  • Consider which kind of AI you're building — Acemoglu's "so-so automation" that replaces workers, or AI that genuinely complements human capability.

If you're a worker:

  • Transition mechanisms matter more than predictions. Demand concrete policy — UBI, retraining, cooperative ownership, public AI access — not just aspirational framing.
  • The "liberation from labor" narrative is worth aspiring to and scrutinizing. The difference between liberation and displacement lies in institutional design.

The Bottom Line

Bougueng's letter attempts something rare and valuable: connecting the technological, economic, social, and philosophical dimensions of AI into a single coherent picture. It doesn't always succeed — the species metaphor needs grounding, the UBI bridge needs company, and the post-labor vision needs institutional architecture. But the ambition is right.

The most important contribution may be the letter's insistence that we think at civilizational scale. Too much AI discourse gets trapped in quarterly earnings cycles or regulatory compliance checklists. Bougueng asks the bigger question: what kind of civilization are we building? And he's right that the answer isn't predetermined.

What the letter needs — and what the broader conversation needs — is more architectural detail. Not just what future we want, but what institutions, policies, ownership structures, and governance frameworks get us there. The ideas exist, scattered across Ostrom's commons governance, Mazzucato's public investment frameworks, Franklin's holistic technology, Benkler's peer production, and Acemoglu's labor-complementing AI. The next step is connecting them as deliberately as Bougueng connects his diagnosis — building a map not just of the problems, but of the paths forward.

The future isn't something that happens to us. It's something we build. Letters like this one help us see the blueprint — even when the blueprint still has gaps that need filling.

Sources

  • Bougueng T., Renaud. "AI and Human Civilization: Strategic Foresight." Maisons Noaré, March 18, 2026. Maisons Noaré
  • Acemoglu, Daron and Simon Johnson. Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity. PublicAffairs, 2023.
  • Acemoglu, Daron. "The Wrong Kind of AI? Artificial Intelligence and the Future of Labour Demand." Cambridge Journal of Regions, Economy and Society 13, no. 1 (2020). Oxford Academic
  • Benkler, Yochai. The Wealth of Networks: How Social Production Transforms Markets and Freedom. Yale University Press, 2006. Publisher
  • Crawford, Kate. Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press, 2021.
  • Franklin, Ursula. The Real World of Technology. House of Anansi Press, 1989. Publisher
  • Korinek, Anton. "The Distributional Effects of AI." Working paper, 2024.
  • Maslow, Abraham. "A Theory of Human Motivation." Psychological Review 50, no. 4 (1943): 370-396.
  • Mazzucato, Mariana. Mission Economy: A Moonshot Guide to Changing Capitalism. Penguin, 2021. Publisher
  • Mokyr, Joel. The Lever of Riches: Technological Creativity and Economic Progress. Oxford University Press, 1990.
  • Aghion, Philippe and Peter Howitt. "A Model of Growth Through Creative Destruction." Econometrica 60, no. 2 (1992): 323-351.
  • Ostrom, Elinor. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press, 1990. Publisher
  • Hinton, Geoffrey. "Two Paths to Intelligence." Talk, NeurIPS 2024.
← Back to all articles
On this page