I went through an opinion piece in the New York times by Evgeny Morazov (NYT Sunday, July 2, 2023). I don’t disagree with the point he wants to make, but there’s a lot of argument being made to support it that is weak, and the thesis itself is weakly laid out. The premise, as I understand it, is that ‘NeoLiberalism’, by which he means something like confidence in technocratic society to achieve positive social benefits, actually fails at that task and hence, confidence in what AGI (Artificial General Intelligence) can accomplish is misplaced. His argument is bound up in pointing out that occasions where he believes past parallels have failed. Central to the piece are: 1) Assertions of failure or more skewed representations of what underlay the failures; 2) An assertion that these failures are sufficient to generalize to AGI; and 3) Assertions that supporters arguments, as he represents them, are weak.

Now, again, I don’t disagree with his conclusion. I don’t know that its right, but it is possible that continual improvement of technology will lead to tools for technocrats that will ultimately not align with broader social benefit. But the points themselves to conclude this in any strong manner aren’t there. He leans heavily, for example, on how Uber has not been transformative in terms of encouraging public walking space and how we travel. Within that argument he ties the hope of Uber to public transportation, but the right connection for now is to Taxis not Subways and buses. He points to how having apps that help individuals track their health has not ended obesity. But is that really the right connection? Who exactly is claiming that my nutrition/exercise app is going to end obesity in society? It’s also an argument that fails to think in terms of counter-factuals.