Distribution

Emad Mostaque, the former CEO of Stability AI, made a counterintuitive argument that I keep thinking about.

His argument is that UBI can become mathematically impossible under mass AI displacement, even when the policy intention is good.

The Math

AI kills jobs at scale. When jobs disappear, spending power disappears. Consumer spending drives GDP. Lower spending means lower tax revenue. Lower tax revenue means you can't fund basic income. Basic income was supposed to be funded by productivity gains. But the people receiving it can't spend the money because there's no economic activity to spend it into.

It's a spiral. Displacement → lower aggregate demand → lower tax revenue → can't fund UBI → people can't spend → lower demand → recession.

Some version of this plays out in every developed economy that tries pure UBI without restructuring its tax base.

The AI Company Problem

AI companies capture much of the benefit from displacement, and their profits do not necessarily remain in the tax base.

They reinvest everything. They raise more capital. They move fast. They also, notoriously, avoid taxes. Profits get sheltered in subsidiaries, paid out as stock, funneled through tax havens.

So you need UBI funded precisely when the companies generating the surplus are structurally motivated to minimize tax liability. Good luck. You've essentially asked the winners to voluntarily fund the losers. Even with aggressive policy, capital's nimble. It will find loopholes.

A production-based alternative

Mostaque's alternative was personal digital currency where individuals "mint money" through self-improvement and community contribution.

The mechanism works like this.

Imagine everyone gets access to a personal AI tool, maybe $1/month subsidy. But the value of that tool compounds based on what you do with it. You use it to learn a skill that gets you a job, your personal currency appreciates. You use it to train others, appreciates. You use it for creative work that others value, appreciates.

You produce value, and the medium of exchange is tied directly to that activity rather than GDP or tax revenue.

This solves several problems simultaneously:

1. Constant economic demand. Everyone has a reason to use the tool because it becomes their economic engine, with no government mandate required.

2. Self-directed value creation. You build an economic position in real time instead of waiting for a job that does not exist.

3. No tax revenue requirement. Participation and contribution mint the value without extracting it from existing economic activity. Funding does not require tax revenue to grow at 3%.

4. Feedback loops work. You get better at something, your currency strengthens, you can access better tools, you get even better. It's upward.

5. Voluntary participation. You join because it can make you richer, faster than the alternatives. The economic reward supplies the incentive.

The Infrastructure

This requires several pieces:

First: AI tools that are useful for skill-building rather than toy chatbots. Language models that can teach programming, medicine or design need to do the job well.

Second: some mechanism to assess value contribution. This is fuzzy. It's not perfect. But communities can evaluate whether you're adding value, through peer review, through measurable outcomes, through contribution tracking.

Third: a currency system minted based on contribution, not borrowed against future tax revenue.

Fourth: integration with existing financial systems so this currency has real purchasing power.

What UBI leaves out

UBI addresses displacement through redistribution and assumes a durable tax base. Extreme AI scenarios put that assumption under pressure.

Personal digital currency starts with production. Everyone has a tool, contribution can be recognised in real time, and participation supports continuing demand.

The model is imperfect and makes no romantic claim for markets. It can survive if GDP growth stalls or traditional employment disappears because it bets on human productivity rather than tax revenue.

The Real Shift

Mostaque was right that UBI misses something fundamental. It treats displacement solely as redistribution and misses the loss of meaning.

When you can't work, you lose purpose. Money without purpose leaves suffering with a better bank account.

Personal digital currency gives everyone a reason to engage, improve and contribute. Each person participates in creating value, and the system can support that participation even if everything else falls apart.

That is the architecture Mostaque argues we should build. The next step is to test its weakest seam: how contribution is assessed without turning community judgement into another centralised gatekeeper.