Redistributed

As AI automates more of material production, a seductive idea keeps resurfacing: what if cryptographic systems — tokens, smart contracts, decentralized autonomous organizations — could route AI-generated value directly to people, bypassing the corporations that currently capture it?

Renaud Bougueng T. articulates the vision clearly: "post-economies centered on authentic human-to-human experiences augmented by AI, potentially powered by novel cryptographic value systems" (Bougueng, 2026). In this framing, AI handles production while new token-based economies emerge around what machines cannot replicate — human creativity, connection, experience. Blockchain becomes the plumbing for a post-labor economy.

It is an elegant idea, and the people building toward it are tackling a real problem. Whether the specific mechanism works is another matter — but the diagnosis deserves to be taken seriously even where the prescription falls short.

The Vision

The argument runs like this. AI will increasingly automate cognitive labor — writing, coding, analysis, design — the way industrial machinery automated physical labor. When that happens, the current economic model breaks. If machines do the work, wages stop flowing, and the mechanism that distributes purchasing power to most people collapses.

Cryptographic systems offer a potential replacement. Instead of employers paying wages for labor, token systems could distribute value based on different criteria: contribution to shared resources, participation in governance, or simply existence (a universal basic token). Smart contracts could encode these distribution rules transparently. DAOs could govern the systems democratically. No corporate intermediary needed.

This is not pure speculation. Pieces of it already exist.

What Already Exists

Proof-of-work and proof-of-stake systems allocate new tokens based on computation (mining) or existing holdings (staking). These are distribution mechanisms, just ones that reward capital and hardware ownership — a point we will return to.

DAOs attempt democratic governance of shared treasuries and resources. As of early 2026, DAOs collectively manage over $30 billion in assets (DeepDAO, 2026). Some, like Gitcoin, fund public goods through quadratic funding mechanisms designed to amplify small contributions.

Token-based creative economies let artists and creators issue their own currencies. Music NFTs on platforms like Sound.xyz give fans direct ownership stakes in songs. Social tokens let communities create their own internal economies. These experiments have been uneven — many collapsed during the 2022-2023 crypto downturn — but the surviving projects demonstrate that peer-to-peer value exchange without intermediaries is technically possible.

DeFi protocols disintermediate financial services: lending, borrowing, trading, and insurance without banks. By early 2026, DeFi protocols hold roughly $150 billion in total value locked, down from speculative peaks but persistent enough to suggest something durable (DeFi Llama, 2026).

Data cooperatives are exploring token-based data ownership, where individuals contribute data to shared pools and receive tokens proportional to their contribution. Projects like Ocean Protocol and Streamr attempt to create markets where data has a price set by its contributors rather than extracted by platforms.

Where Crypto Could Work for AI Redistribution

Some applications have genuine potential.

Training data provenance and compensation. Blockchain's immutable ledger is actually well-suited for tracking which data trained which model. If you could record data lineage on-chain, you could automate royalty payments when a model trained on your work generates revenue. This addresses the core complaint of artists, writers, and creators whose work feeds AI systems without compensation. See Who Owns the Training Data?.

Decentralized compute networks. Projects like Akash Network and Render Network use token incentives to aggregate underutilized computing resources — idle GPUs in homes and offices — into shared networks. Providers earn tokens, users get cheaper compute. This could meaningfully broaden access to AI inference and fine-tuning, currently bottlenecked through a few cloud providers. See The API Economy.

Tokenized model ownership. A DAO could collectively fund model training and distribute ownership tokens to contributors — data providers, compute donors, researchers. Revenue from the model's API would flow to token holders. This is cooperative ownership with cryptographic enforcement, and several projects are attempting versions of it.

Universal basic tokens. Instead of universal basic income funded by taxation, token systems could mint new tokens and distribute them equally. Worldcoin, despite its considerable problems (iris-scanning in the Global South raises questions Langdon Winner would recognize about who designs the artifacts), represents an attempt at this model. The redistribution question is whether such systems can maintain value while distributing broadly — a challenge that has defeated most attempts so far. See Universal Basic Income and AI: The Debate Gets Real.

Where It Fails

Here is where the vision collides with a decade of evidence.

Crypto wealth concentration mirrors — and in some cases exceeds — traditional wealth concentration. The top 1% of Bitcoin addresses hold approximately 30% of all Bitcoin. Ethereum's distribution is similarly skewed. As Vitalik Buterin himself has acknowledged, token distributions tend to concentrate over time as sophisticated actors accumulate holdings from less sophisticated ones. If the mechanism meant to redistribute value reproduces the same concentration, it has not solved the problem. It has rebranded it.

Token systems create speculative assets, not stable economies. Most tokens function as investments, not currencies. Their value fluctuates wildly, which makes them poor foundations for stable economic participation. The people who can tolerate volatility are the people who can afford to lose money — which is to say, the people who least need redistribution.

DAO governance struggles with the same problems as all governance, plus new ones. Voter apathy is endemic: most DAOs see participation rates below 10% on proposals. Whale dominance — where large token holders control outcomes — reproduces plutocracy with cryptographic characteristics. Coordination costs are high. As Elinor Ostrom demonstrated, commons governance works through specific institutional arrangements built over time (Ostrom, 1990). DAOs have generally skipped this institutional work in favor of elegant code, and it shows.

Smart contracts encode their creators' values. A smart contract executes rules written by someone. The question "whose values does this encode?" applies with the same force as it does to any AI system. As Winner argued in "Do Artifacts Have Politics?", technical systems embed political choices (Winner, 1980). A smart contract that distributes tokens based on compute contribution rewards those with hardware. One based on data contribution rewards those with data worth contributing. The distribution rule is a political decision dressed in mathematical notation.

Energy costs remain significant. Ethereum's move to proof-of-stake reduced its energy consumption by over 99%, but the broader crypto ecosystem — including Bitcoin, which shows no signs of transitioning — still consumes more electricity than many countries. For a technology positioned as part of a sustainable post-economy, this is a material contradiction. See the environmental cost of AI.

A decade of results. Web3 has been promising redistribution since at least 2015. The observable outcomes: a small number of people became very wealthy, a larger number lost money, and the populations most in need of economic redistribution — those without hardware, connectivity, technical literacy, or spare capital to risk — remain almost entirely excluded. The overlap between "people who benefit from crypto" and "people who need redistribution" is distressingly small. See The Digital Divide in AI.

Who Benefits, Who Pays

The redistribution question cuts through the rhetoric.

Who benefits from cryptographic post-economies? Early adopters. The technically literate. Those with capital to invest in tokens, hardware to participate in networks, and time to navigate complex protocols. Venture capitalists who fund crypto infrastructure. Exchanges that take fees on every transaction.

Who pays? Late adopters who buy at inflated prices. People in developing countries whose electricity subsidizes mining operations. Communities whose attention and data feed platforms that issue tokens they cannot meaningfully use. And, most quietly, everyone who invests hope in a technological solution while the window for institutional solutions narrows.

This is not a new pattern. As Virginia Eubanks documents in Automating Inequality, technological systems positioned as neutral or liberatory often reproduce existing hierarchies through new mechanisms (Eubanks, 2018). Cryptographic systems are not exempt from this dynamic because they are decentralized. Decentralization distributes control; it does not automatically distribute power.

What Would Actually Work

If the goal is redistributing AI-generated value, the most promising approaches are less elegant but better evidenced.

Commons governance. Ostrom's framework for managing shared resources — clear boundaries, proportional costs and benefits, collective decision-making, graduated sanctions — has been validated across hundreds of real-world cases. Applying these principles to AI resources (training data, compute, models) does not require blockchain. It requires institutional design. See AI Cooperatives: Alternatives to Corporate Models.

Public infrastructure. Government-funded compute facilities, publicly maintained open models, regulated data trusts. The boring machinery of public provision, which has successfully distributed electricity, water, postal services, and telecommunications to billions of people. None of these required token systems.

Regulated markets. Taxation of AI-generated profits, mandatory data licensing fees, antitrust enforcement against compute monopolies. The mechanisms exist. The political will is the variable.

Hybrid approaches. Crypto technology is not useless — provenance tracking, micropayments, and programmable money have genuine applications. The best projects in the space recognise this: Gitcoin's quadratic funding has distributed over $50 million to public goods, demonstrating that crypto mechanisms can serve redistributive ends when deliberately designed to do so. But these tools work best as components within governed systems, not as replacements for governance itself. A blockchain that tracks data provenance is more useful inside a regulatory framework that requires compensation than floating in a libertarian vacuum.

The Infrastructure Question

Bougueng's vision of post-economies is worth taking seriously as a diagnosis. AI is transforming how value gets created, and current institutions are not distributing the gains broadly. The question is whether the prescription — cryptographic value systems — matches the disease.

The evidence so far suggests that cryptographic mechanisms are good at creating new markets and poor at ensuring those markets serve broad populations. They are infrastructure for exchange, not infrastructure for equity. And the difference matters.

Redistribution is not a technical problem. It is a political one. The technology that best supports it is whatever technology operates within institutions designed to distribute power broadly — whether that technology runs on blockchain or a spreadsheet.

Sources

Bougueng T., Renaud. AI and Human Civilization: Strategic Foresight for Our Shared Future. 2026.
Ostrom, Elinor. Governing the Commons. Cambridge University Press, 1990. Publisher
Winner, Langdon. "Do Artifacts Have Politics?" Daedalus 109, no. 1 (1980): 121-136. JSTOR
Eubanks, Virginia. Automating Inequality. St. Martin's Press, 2018. Publisher
Crawford, Kate. Atlas of AI. Yale University Press, 2021. Publisher
DeepDAO. "Organizations Overview." 2026. DeepDAO
DeFi Llama. "Total Value Locked." 2026. DeFi Llama
Chainalysis. "The 2024 Geography of Cryptocurrency Report." 2024. Chainalysis

General

Cryptographic Post-Economies: Can Token Systems Redistribute AI Value?