The United States may be in danger of creating a ‘tokenocracy’ in which justice, opportunity, and self-governance flow toward those with the largest AI budgets. | Faculty fellow @KevinTFrazier for @techpolicypress

A competitive AI ecosystem is essential the technology realizing its potential to expand opportunity and increase access to information. The good news is that several labs claim they support this vision. Cato’s @KevinTFrazier points out that’s it’s still unclear whether they’ll act on it.

Huge congrats to all the academics heading to what I think we ought to call the "Silicon Tower." AND, boy, do we need to have a real conversation about the brain drain and the current state of our AI talent pipeline. There's a huge rule of law problem when only the labs and the government understand what's truly happening at the frontier (it's made even worse when the govt starts relying on secret frameworks). (h/t @_NathanCalvin, @jasminewsun, & others who have been teeting about this). Regardless of your preferred regulatory framework (or lack thereof), it's imperative that we have a steady stock of independent AI researchers and that the inflow exceeds the outflow. That's not the case right now. We need much greater general investment in traditional literacy and AI literacy to help more everyday Americans participate in AI governance conversations (it's not great when a lot of the policy discourse occurs on a platform that many folks avoid). AND, we need a specific, immediate investment--private & public--in AI expertise, defined broadly. We cannot hope to hold the government accountable nor to constrain the power of the labs if we do not have a vibrant independent AI ecosystem. I wrote about this on Sunday. I flagged the excellent research by Zahra Meghani, who warned, "[W]hen regulatory agencies choose to rely primarily or entirely on product sponsors for data, analyses, and evaluations, they in effect voluntarily subject themselves to an epistemic monopoly. In doing that, they create conditions for their own domination by those actors." I will continue to be writing about this. For now, I want to share a similar instance in which we've recognized a talent shortage and rapidly mobilized to address it. Engineering, Science, and Management War Training - 1940 to 1945: In under five years, 1.5 million Americans received "short, intensive college-level courses designed to prepare for technical and scientific work in war industries. these courses were conducted by colleges and universities under the sponsorship of the U.S. Office of Education." Imagine a similar focus today. If you're looking to be part of the solution, here's a question to keep asking and working on: What discrete competencies must an IVO possess, how many people with each competency will be required in 6, 12, and 24 months, and which existing institutions could convert people into those occupational categories in weeks or months?