> Tao and others in that group have been strongly and publicly pro AI from the start
Unfortunately being "pro AI" means relinquishing any control over what the AI, or more importantly the company running it, might be doing.
Unfortunately being "pro AI" means relinquishing any control over what the AI, or more importantly the company running it, might be doing.
We've relinquished control over just about everything we use or consume. We can't compete with larger enterprises for production of food, clothing, machinery, medicine, energy, services. Mathematics is just the latest thing to be industrialized.
What keeps large companies under control is competition with other large companies. This competition causes the surplus value they produce to flow to consumers, not be hoarded via monopoly prices. Do we see strong moats that are going to cause monopoly in AI? I don't see it, and in particular I don't see it persisting if it exists transiently.
> Do we see strong moats that are going to cause monopoly in AI?
Ownership of the capital assets used to train and inference new models. Yes, we may end up with more than one firm. But as we see with big tech today, a small number of fantastically wealthy firms in "competition" does not an open market make.
Somehow you have to argue either that society itself is bad, or that math is somehow different from all these other human activities.
I think the obvious fact that people prefer to live in places with large commercial organizations shows they don't really care about that, at least to the point of foregoing the benefits these organizations bring.
The current model of society in the US is undeniably bad. The idea of “checks and balances”, taught in high school civics, has essentially completely disappeared.
But that’s a medium-term problem, which will correct itself. People who think they can eliminate checks on themselves eventually find out that they’re wrong.
Computers and computer programs are tools. Humans always remain sovereign over their tools.