Samizdata

I get the daily posts from the Law & Liberty blog, and this struck me as interesting, because of the preamble:

Dozens of start-ups now offer Artificial Intelligence tools to help businesses set market prices. Assuming unlimited computing power to run such models and comprehensive data sets to train them, can AI replicate the way human actors make decisions in the marketplace? Socialists have argued for more than a century that enlightened bureaucrats can set prices as well as the myriad of private actors in the marketplace. Ludwig von Mises offered a celebrated refutation of the socialist case. Does the vast computing power behind Large Language Models give new life to the socialist argument? The answer is no, but Mises’ argument needs to be updated and sharpened.

The author of the article, David P Goldman, goes on to explain the problem. As the article is free to access, I won’t reproduce other paragraphs here apart from the two final ones:

AI can’t replace the innovative creativity of entrepreneurs. On the contrary: AI itself is an innovation whose outcome is uncertain. Some applications (replacing human beings on corporate help desks, for example) may turn out to be trivial; others, for example devising new pharmaceuticals, may be revolutionary. Only in an imaginary world in which no innovation occurs could we envision an AI-driven marketplace.

Artificial Intelligence isn’t intelligence in the first place. It can replicate the lower-order functions of the human mind, the sorting and categorizing faculty, and perform such operations much faster than humans. But it cannot reproduce the higher-order functions of the mind—what Immanuel Kant called Vernunft (roughly, critical reason) as opposed to Verstand (usually translated as “understanding”). It can mine data from past experience, but it can’t stand at a distance from experience and ask, “What if we did things differently?” Freedom is the freedom to create, and that is what free societies must preserve.

This seems right to me. I think AI is going to produce marvels, but I don’t see it removing the need for boldness, risk-taking and ability that all great businessmen have to “look around corners”. To ome extent I am a techno-optimist, as the likes of Marc Andreessen, the US venture capitalist, is. But I am not, I hope, Panglossian, or the opposite of a perma-doomster, either.

It is also interesting to consider how governments, for example, might seize the idea that AI makes it possible to co-ordinate human activity in ways that eliminate all that pesky free market exchange and messy entrepreneurship. This line of thinking resembles the view of certain science fiction writers who tried to imagine a post-scarcity world. (Science fiction often contains lots of economics, as this article by Rick Liebling shows.) Eliminate the idea of scarcity, so the argument runs, and then the underlying foundation of economics – “the study of scarce resources that have alternative uses” – falls away. It is easy to see the utopian attractions if you like to mould humanity to your will. I mean, what could go wrong?

Eliminate scarcity, then who needs enforceable property rights and rules about “mine and thine”?

In a post-scarcity world, where will the sense of urgency come – the sense of adventure, that drives great businessmen to create and innovate to push back against such scarcity? (This is also the fear that some might have of universal basic income – creating a world of indolent trustafarians who, like a couch potato, suffer muscle loss and mental decline because they don’t have to work or struggle to build anything.)

Karl Marx dreamed of a post-scarcity world – that seems the logical end-point of his communist utopia, to the extent he fleshed it out at all. (The irony being that his ideas helped inspire some of the greatest Man-made famines and loss of life in recorded history, in part because of the failure to understand the importance of property, prices and incentives.)

I am sure that some of this post-scarcity thinking might be encouraged by AI. But then again, AI uses a lot of electricity, and even without the distractions of Net Zero (no laughing in the class, people), producing the power necessary for modern high-potency computing requires a lot of stuff. And mention of science fiction reminds me of the “There Ain’t No Such Thing As a Free Lunch” that came from Robert A Heinlein, and later taken up by Professor Milton Friedman.