AGI Is Here and Will Never Be

Claude 3 Opus and AGI - Marginal REVOLUTION

As many MR readers will know, I don’t think the concept of AGI is especially well-defined. Can the thing dribble a basketball “with smarts”? Probably not. Then its intelligence isn’t general. You might think that kind of intelligence “doesn’t matter,” and maybe I agree, but that is begging the question. It is easier and cleaner […]

Marginal REVOLUTIONTyler Cowen

Tyler Cowen makes a simple point regarding the "AGI" terminology debate:

I would say this: there is yet another definition of AGI, a historical one. Five years ago, if people had seen Claude 3 Opus, would they have thought we had AGI? Just as a descriptive matter, I think the answer to that question is yes, and better yet someone on Twitter suggested more or less the same. In that sense we have AGI right now.

I also think it's quite possible we never get to a place where everyone agrees that AGI is here. There are too many possible definitions and the goal posts will keep moving. I also agree with the point that many people have various incentives and biases so as to likely never be fully satisfied that such technology is here.

Incredible technology is already here. Even more incredible technology will be here tomorrow. But it's never going to cross some obvious line in the sand.