SEBASTOPOL, CALIFORNIA -

Tim O’Reilly says more

You have called the generative artificial intelligence [GenAI] “bubble” a “productive” one because the tech is being applied not only in the “betting economy,” but also in the “operating economy,” with “massive amounts of capital” being “used to build the infrastructure of the future.” But as you write in your 2017 book WTF?: What’s the Future and Why It’s Up to Us, in “business and in technology, we often fail to see clearly what is ahead because we are navigating using old maps and sometimes even bad maps.” How, then, can the world ensure that the AI infrastructure built today proves broadly beneficial tomorrow?

Great question. The whole point of a bubble is that it’s a wild exploration of new possibilities - blank spaces on the existing map. But understanding the base layers of the map that still constrain our vision is important. 

One unexamined base layer that is impeding the development of broadly beneficial AI is the imperative of modern capitalism - especially in the United States - which places shareholder returns above all other considerations. Until we confront this limiting idea, AI governance is likely to go the way of the ‘environmental, social, and governance’ movement - at best being paid lip service, and at worst being actively ignored. I wrote about this last year in a piece called The Alignment Problem Is Not New. 

There is a lot of focus on the risk