Equitable distribution

That’s the grand bargain: build it, but make it work for everyone. That means discovery, progress, and growth alongside safety, shared prosperity, and the protection of—the exaltation of—what is distinctly human.

It’s the right impulse and the right moment. Investment is high (so there are resources to spend on concessions) and political counterpressure is building. I’ve said that stopping AI development is unrealistic, but credit where it’s due: forceful opposition creates the conditions where strong, workable ideas can take hold.

The question is: what form should the bargain take?

There are versions that are attempts to lock in the present, or even the past. Bill Gates’s proposal to create employment preserves for humans—that is, banning AI from doing jobs it does better so that humans will have something to do—comes to mind. Other ideas, like token taxes or profit-sharing with AI companies, assume that constructs like tokens or specific AI companies will endure as the way the economy is set up. We are so early in this era, and we have no idea what the shape of AI will be in five or ten years. The companies that seem unassailably dominant today could be afterthoughts by then, like DEC, Compaq, Yahoo, or BlackBerry before them, and the companies that realize more of the value might not fit the way we set up legislation.

These proposals are brittle, either because they try to prevent rather than adapt to the changes AI might deliver, or because at their core they assume that what we understand of AI and its market today will be true in the future.

Then there are things we can do that are valuable no matter what direction things go. Expanded technical literacy for children; we may not know where technology is headed, but plainly we are not teaching our kids even the rudiments that they’d need to know in any future. Changes to higher education to help more adults prepare for the work that truly demands human work; we may not know what work is yet, but a system that demands years out of the workforce and tens or even hundreds of thousands of dollars in debt is not designed for this era. Broad-based tax reform to make sure AI-driven growth expands public capacity, regardless of where profits pool, and deployment of that capacity in the service of taking better care of people, in the many forms that can take.

And, a revolution in the relationship between public and private, so that all people not only benefit from but have a voice in creating the technology.

Those are ideas that do not fight with the transition, and recognize that we can’t know how it will play out, and therefore can accomplish what we need. That’s the shape of the bargain we need.