AI and Robots Could Create Abundance. Who Gets a Share?

Elon Musk now has a more concrete answer to the question of how people might share in the wealth he expects machines to create: give them some ownership. On October 19, he said he would give one of his own SpaceX shares to every child in America in 2027.

A share is not a salary. But the promise puts something tangible behind a much larger idea: people should be able to benefit from increasingly productive machines without first getting a job at the company that owns them.

Musk is not the only one exploring that territory. Anthropic has pledged equity for public benefit. OpenAI has proposed a public wealth fund, and its foundation has committed money to developing new economic institutions. These efforts differ sharply in who would own the assets and decide what happens to them. Together, they suggest an important shift: the abundance debate is beginning to acquire financial architecture.

The early signals behind the singularity debate

The technological idea is a feedback loop. AI helps researchers develop better AI; the improved systems help with more of the next round of research. If that loop becomes powerful enough, progress could accelerate beyond the pace of a mainly human research workforce. This is one of the mechanisms behind what is often called the technological singularity.

A September 28 paper by researchers including Geoffrey Hinton, Yoshua Bengio and Jack Clark examines that possibility. It describes AI taking on substantial parts of AI research and development, while emphasizing that an intelligence explosion remains uncertain. The authors say productivity gains from this automation had not yet reached the threshold required to trigger one. Computing resources, difficult tasks and real-world experiments can still slow the loop.

The economic significance does not depend on declaring that a threshold has already been crossed. If the tools used to discover and build things become much more capable, the capacity to produce useful goods and services can change with them.

Robotics makes that possibility physical. In a September account of its factory pilots, BMW reported that Figure 02 had moved more than 90,000 components during a 2025 deployment that supported production of over 30,000 X3 vehicles. Its job was to position sheet-metal parts for welding, not build entire cars alone. This was a bounded industrial task, but it was real production rather than a stage demonstration.

That is the connection worth watching: more capable software, machines that can act on the physical world, and new attempts to give people a stake in the value those systems may create. The first two concern production. The third concerns who gets to participate.

Musk is moving from an income vision to an ownership pledge

Musk has long described the destination in sweeping terms. In a July 2 post, he linked AI and robots to universally high incomes and work becoming optional. That is his forecast, not a description of today's economy.

His new share pledge is narrower and more specific. It names an asset, an intended population and a year. It is also a personal promise involving his existing SpaceX holdings, not a SpaceX program distributing robot profits. The October 19 post leaves eligibility details open.

There is a precedent within the same company. In July, Gwynne Shotwell and her husband pledged personal SpaceX shares for more than two million children through Trump Accounts. Musk has also responded positively to using that account route. The Treasury has established a pathway for philanthropic stock contributions, although a named pathway and an announced gift are not proof of completed transfers.

What makes this relevant to the abundance argument is ownership. A person who receives a share can participate in the company's future value rather than relying solely on a wage from it. A childhood investment account is a long-term stake, however, not money a household can immediately use to replace a lost paycheck.

A company fund and a public fund are different answers

Anthropic is taking a pooled approach. Its October 18 announcement commits an initial 5% of its shares to a Public Benefit Fund, subject to stockholder approval. The fund would be wholly owned by Anthropic and could support nonprofits or make distributions to particular populations.

The attraction is that the stake can become more valuable as the business grows. The distinction is that people do not automatically receive shares or an individual claim on payments. The fund remains the intermediary. Our separate explainer on Anthropic's proposal examines that structure in detail.

OpenAI's April policy paper proposes a different kind of pool: a Public Wealth Fund that could give citizens, including people who own no investments, a stake in AI-driven growth. It could hold a diversified portfolio of AI companies and businesses using AI, with returns potentially distributed to citizens. Policymakers and companies would still need to decide how to finance it. This is a policy proposal, not an operating fund.

The paper also widens the definition of a dividend. Workers might receive better benefits or some productivity gains as time: it proposes trials of a 32-hour, four-day week without reduced pay, while maintaining output and service levels. Shared prosperity need not arrive only as a check.

The OpenAI Foundation has separately committed an initial $250 million to its Economic Futures program. Announced on May 27, the commitment will support grants, partnerships and direct work on economic security, including public wealth funds and durable claims on growth. That budget is for developing and testing institutions, not a $250 million pot already being paid to citizens. Still, putting resources behind institutional design is a more consequential step than predicting abundance and assuming distribution will take care of itself.

What would make the promise real?

The common idea is bigger than compensation for losing a job. It is that people could participate in an economy as owners or beneficiaries even when they are not needed to perform every task that creates its wealth. That could widen opportunity well before work becomes optional, if it ever does.

But different routes solve different problems. Cheaper goods can make an existing income go further. A shorter week gives people time. Cash can cover immediate needs. An ownership stake can grow over years. None is a complete substitute for the others, and a gift dependent on a donor is different from a lasting public entitlement.

The next meaningful evidence will therefore be practical: shares credited to real recipients, funds with published rules and resources, or workplace gains that appear in pay, benefits and time off. Those are the steps that turn a compelling future into something people can plan around. Our earlier look at ownership and public agency after AGI asks a further question: whether receiving a benefit also gives people a voice in the system delivering it.

The encouraging development is that the builders of these technologies are beginning to confront that second half of the problem. More powerful AI and more capable robots may expand what society can produce. The emerging ownership plans ask how someone who did not finance the machines, build them or work for their owners could still have a claim on the prosperity they create.