Economics
When the Work Changes Hands
For two hundred years, a job did two things at once. It got work done, and it put money in a person’s pocket, money that flowed back out into the world as rent, groceries, a kid’s shoes, someone else’s wages. Work was never only production. It was how value circulated, how the prosperity a company created found its way back to the people who buy what companies sell.
Artificial intelligence is very good at the first job and does nothing for the second. When a system takes over the work, the work still gets done—often better, often cheaper. But the circulation stops. The paycheck that used to ripple outward through a dozen other lives simply ends. Multiply that across an economy and you don’t get a leaner version of the world we have. You get one where the value still gets made but no longer reaches the people who keep the whole thing moving.
This is not a reason to fear the technology. BEACON is for this technology, it can lift burdens humanity has carried for millennia. It is a reason to be honest about a gap the technology opens and cannot close on its own: the people whose work is displaced are also the customers, the neighbors, the demand that every business ultimately depends on. A company that automates away its workforce wins the quarter and, eventually, automates away its own market. What’s good for one balance sheet this year can hollow out the ground everyone stands on.
So the question isn’t whether AI should be allowed to do the work. It should.
The question is how the value it creates gets back into circulation so that a transformation this large lifts people instead of leaving them behind. That’s solvable. It does not require slowing the technology or punishing the companies building it. It requires making sure that when the work changes hands, the people whose hands it left aren’t simply dropped.
That’s the work BEACON exists to do. Not to stand in the way of what’s coming, to make sure that when it comes, nobody gets left out of it.
THE PROPOSAL: A TRANSITIONAL BRIDGE
The honest picture of AI and work, as of now, is not an apocalypse—and BEACON won’t pretend it is. Aggregate employment remains relatively stable. What’s happening is subtler and, for the people inside it, more dangerous than a headline number reveals: displacement is real, it’s concentrated on specific people, and leading projections expect millions of U.S. workers to need new occupations over the coming decade. The destroyed jobs and the created jobs are not the same jobs, do not require the same skills, do not pay the same wages, and are not in the same places. A displaced administrative worker in one city does not automatically become an AI specialist in another. The gap between those two realities is where the genuine human cost lives.
BEACON’s first policy framework is built to bridge that gap, not to replace all wages forever, but to carry displaced people through the transition with dignity intact.
Who it reaches, and why we don’t ask “did AI do it?”
The cleanest cases of AI’s harm are invisible to any system that demands proof. Consider the graduate who did everything right and walks into a labor market where the entry-level career they trained for no longer hires, not displaced from a job, but pre-empted from ever holding one. There is no layoff notice to point to, no robot’s fingerprints to document. To demand that such a person prove AI caused their predicament is to design a system that systematically excludes the people it should most protect.
So BEACON’s framework does not gate eligibility on proving AI-causation. That test would be both a bureaucratic nightmare and an open invitation to fraud. Instead, eligibility is based on objective circumstance, sustained inability to find work beyond a defined window. Measured through data that already exists, requiring no story that could be faked because no story is solicited. The argument that AI is reshaping the labor market is the policy’s justification. It is not a hurdle placed in front of a frightened person who simply needs to keep their apartment.
How long it lasts, and what it leaves in place
The bridge lasts as long as the transition does for each person. It continues while someone retrains or searches for work, and it tapers gradually as their earnings return, designed so that taking a job always leaves people better off than staying on the bridge. For people whose work does not come back, the bridge hands off to the broader arrangement described below rather than ending at a cliff. That handoff is deliberately left open. As the wider economy changes, the bridge should adapt to fit whatever long-term solution society ultimately chooses.
It also works with the protections people already have, not against them. Unemployment insurance still applies, and the bridge tops it up to the livable floor. And BEACON proposes that the bridge not count as income for Medicaid or other health coverage, and that time spent retraining or searching for work satisfy coverage work requirements, so no one loses their health care by accepting it. A floor that costs someone their health coverage would not be livable.
What it pays, and what it costs
A transitional income has to be livable, or it isn’t a bridge—it’s a gesture. The figures floated in past UBI debates, around $1,000 a month, were never enough to survive on; they were political numbers, designed to seem affordable while delivering too little to matter. BEACON’s working floor is a genuinely livable range: roughly $2,000 to $2,500 per month—enough to cover housing, food, and essentials while a person retrains, searches, or transitions.
The scale is deliberately modest and grounded in conservative assumptions about how many people the transition displaces above the normal baseline. These are BEACON’s own working estimates, presented as ranges rather than false precision, and we will revise them openly as the data matures. The figures below model the program at its transitional peak, not as a permanent steady state.
The money circulates — it does not disappear. Every dollar delivered to a displaced household typically returns more than a dollar in economic activity, especially when unemployment is high, because people at the edge spend immediately, on rent, groceries, local services. The bridge is not a cost extracted from the economy; it is demand kept in circulation. A company that contributes to it is, in part, funding the continued solvency of its own future customers. This is the core of the argument: what looks like a transfer is actually the economy refusing to hollow out its own foundation.
The destination: from bridge to dividend
The bridge is not the goal. It is what carries us to the goal. As AI-driven abundance matures and the economy genuinely produces more with less, the targeted transitional model should give way to something broader and simpler: a Universal Respectable Income, not a minimal subsistence check, but a dignified share in the abundance the technology creates, paid to all, with the gatekeeping dissolved because scarcity itself has dissolved. In that world, no one is left on the street, without healthcare, education, or the basic respectability of a decent life, and anyone who wishes to earn more remains entirely free to do so. The bridge is targeted because today’s abundance does not yet reach everyone. The destination is universal because, eventually, it can be.
There is already a working model for this. Since 1982, Alaska has paid every eligible resident a yearly dividend from the state’s oil wealth, on the principle that a resource held in common should benefit the people who hold it. It has outlasted governors of both parties and remains broadly popular, even as Alaskans debate its size each year. AI differs in one important way: private companies built it, but they built it on the work and knowledge of countless people. The principle carries over. When value is built on what everyone contributed, everyone should share in it.
We are not claiming that world is here. We are building the path to it, one defensible step at a time.
| Livable monthly payment | $2,000 – $2,500 / month ($24,000 – $30,000 / year), including any unemployment insurance |
| Eligible population at peak | ~4 – 12 million (displaced workers + pre-empted new entrants, above baseline) |
| Annual program cost at peak | ~$100 billion – $360 billion / year (less where unemployment insurance already covers part) |
| Health coverage | Medicaid eligibility protected; adds up to ~$35 – $105 billion / year, much of it spent today regardless |
| Duration | Tapers as earnings return; adapts as the long-term model takes shape |
| Economic return per dollar | Est. $0.80 – $2.20 per $1 for transfers (CBO); ~$1.50 for food assistance in a weak economy (USDA). Highest when unemployment is high |
| Funding source | Bridge-sized levy on captured AI-generated value |
| For scale | Even with health coverage (~$130 – $465 billion total), a fraction of universal UBI models (~$2 – 4 trillion), and well below the ~$2.6 trillion the federal government spends each year on Social Security and Medicare alone |
Figures are BEACON’s conservative working estimates, presented as ranges and revised openly as data matures. Modeled at transitional peak, not steady state.
These figures are illustrative, not precise projections. The mechanism is capable of generating meaningful support at the scope of the problem it addresses.
Sources for the figures above: 2026 federal poverty guidelines · BLS employment data · McKinsey Global Institute on AI and U.S. jobs (via Fortune) · CBO multiplier ranges (via CRFB) · USDA on the SNAP multiplier · MACPAC Medicaid spending per enrollee · Federal spending, FY2025 (via AAF) · OpenResearch cash study · Tax Foundation on sales-based apportionment · Alaska Permanent Fund Dividend Division
Anticipated Objections
“Companies will move AI operations offshore.”
A contribution tied to value generated and customers served, not to corporate headquarters, is hard to escape by relocating. Most U.S. states already tax corporate income this way, by where a company’s customers are rather than where it is based. A company serving customers in a market participates in that market’s obligations regardless of where it’s domiciled.
“This will slow AI development and innovation.”
The contribution is sized to bridge a transition, not to penalize deployment. Companies keep the rest of what AI earns them. And the real threat to AI’s future isn’t a modest, circulating levy; it’s the social and economic instability of mass displacement handled badly. A transition that leaves people behind produces the backlash, regulation, and unrest that genuinely would halt the technology. Funding the bridge protects the conditions under which AI can keep advancing.
“Isn’t this just a tax on technology?”
No—it’s a mechanism for keeping demand in circulation. When AI displaces a worker, the purchasing power that worker carried into the economy disappears, and a chunk of every company’s customer base disappears with it. The contribution restores that circulation, typically returning more than a dollar in economic activity for every dollar delivered. It isn’t a penalty on building things. It’s how the economy avoids hollowing out the demand that makes building things profitable.
“Won’t a basic income reduce the incentive to work?”
The honest answer is that the effect is real but modest. In one of the largest U.S. studies of cash payments, people receiving $1,000 a month for three years worked about 1.3 fewer hours a week than a comparison group. Benefits tied to being out of work can also modestly lengthen job searches, which is why the bridge tapers as earnings return rather than stopping all at once, and is designed so that taking a job always pays. It gives people the footing to retrain, search properly, or take an entrepreneurial risk. The transitional bridge is a floor to stand on, not a ceiling to settle for, and anyone who wishes to earn beyond it remains entirely free to do so.
Implementation Roadmap
Phase 1. Years 1–2: Foundation and coalition. Engage labor unions, worker-advocacy organizations, and displaced-worker communities. Build relationships with economists and policy researchers to pressure-test and refine the model. Develop model legislation for the transitional bridge. Establish BEACON as a credible, grounded voice in the automation-policy conversation.
Phase 2. Years 2–3: Legislative advocacy. Introduce model legislation in the states most open to early action, and build support across the political spectrum, framing the bridge not as a partisan program but as shared economic self-interest. Engage forward-thinking technology companies as partners rather than opponents, since a stable transition serves them too.
Phase 3. Years 3–5: Pilots and national advocacy. Launch pilot programs in early-adopting jurisdictions. Document outcomes rigorously and build the evidence base for broader implementation. Begin developing the parallel frameworks for recognizing and protecting emerging intelligence, so that preparedness keeps pace with capability. Scale toward national and international partnership.
The question before policymakers is not whether AI will displace workers. It will. The question is whether the response will be managed and equitable, or chaotic and destabilizing.