Artificial Intelligence and the Man of System

In this episode, host Russ Roberts welcomes Zvi Mowshowitz to discuss the merits of AI, causes for optimism and concern, AI as a ‘man of system’, and technology as a last bastion of the freedom to innovate.

Artificial intelligence is still in its infancy, but learning to walk quickly. What was once just some chatbots has become a generative AI revolution, redefining the way we communicate, learn, and do business, and it’s just getting started. AI will consistently get smarter, but “smart” in what sense? In this episode, Mowshowitz makes the point that Google  isn’t “smart,” it just has much more access to information, so is generative AI different? Mowshowitz says yes, but it’s complicated. Generally, “smart” means a high level of comprehension, problem-solving, and creative ingenuity. For some questions, chatbots do indeed show comprehension, but Mowshowitz clarifies, that also has a lot to do with the way questions are asked. Roberts adds that he doesn’t know if Chat GPT can think of metaphors that change people’s conceptions of the world, that carry powerful intellectual creativity. 

Mowshowitz is a short-term optimist. He puts concerns such as employment destruction to bed, arguing there are many jobs waiting under the surface. However, Mowshowitz distinguishes between short-term, but present and visible concerns such as misinformation and confusion over deepfakes and the long-term existential questions casting doubt over a prosperous partnership with AI. One such question surrounds creative intelligence itself. How ought AI be maintained? Is it even possible?

On the other hand, a key point for AI optimism lies within the dial of progress. The dial of progress, as outlined by Mowshowitz on his Substack, Don’t Worry About the Vase, is a metaphorical dial displaying a collective decision on whether societies allow individuals to freely innovate and take risks, or require permission to be given and progress to be observed. In Mowshowitz’s eyes, AI is one of the few places that can shoot the dial up. Mowshowitz sees a scourge of red tape in various industries handcuffing prospective innovators, hence the flocking to AI, crypto, and CIS in general. In a sense, AI is a last bastion of opportunity for innovation, for ingenuity, for chaotic dynamism, and potential for rapid progress.

…over the years we’ve moved from a United States that was very much on the, ‘You go out there and there’s an open field and you do more or less whatever you want to do as long as you don’t harm someone else or someone else’s property,’ to a world in which vast majorities of the economic system require detailed permissions that are subject to very detailed regulations that make it very, very hard to innovate and improve. And, I strongly agree with Andreessen and Cowen, and I think you and many other people, that this is very much holding us back. This is making us much less wealthy. This is making us much worse off. And that we would be much better off if we loosen the reigns.”

However, Mowshowitz is not a long-term AI optimist. On the contrary, he argues many optimist models don’t have gears, they’re too simple. To Mowshowitz, these models lack structure, coherence, and logical carryover. Not enough time is spent investigating the inner workings of the world within the model.

“When you have AI in the brain, everything is in some way a metaphor for the problem. Marx writes this huge thing about how capitalism is terrible; we’re going to overthrow it; we’re going to create this Communist utopia. And then, he writes five pages that are completely vague about what the Communist utopia is going to be. We have many other people who do similar things. AI is another example of this where a lot of people are saying, ‘We’re going to build this amazing AI system that’s going to have all these capabilities and then we’re going to have this Brave New World where everything is going to be awesome for us humans and we’re going to live great lives.’ And then, they spend one paragraph trying to explain, ‘What are the dynamics of that world?’ Like, what are the incentives? Why is this system at equilibrium? Why do the humans survive over the long run given the incentives that are inherent in all the dynamics involved?”

Roberts adds fuel to the pessimism, challenging the argument that AI will run amuck. Roberts suggestst AI will embody the goals of its creators, regulators, and users. In his words, the driverless car doesn’t aspire, the more likely concern is human beings using AI to perpetuate harm. He then points to Adam Smith’s man of system. The man of system believes he can solve problems within society by orchestrating individuals just like moving the pieces on a chessboard. However, what he doesn’t understand is the pieces all operate via their own principle of motion completely separate from the will of the man of system. Roberts believes AI faces the same problem, there are clear limits on AI’s ability to impact the world through control, namely a lack of information, particularly when it comes to subjective information which lies in the minds of individuals themselves.

Mowshowitz goes on to state that the man of system is a human being, with a limited ability to access, store, process, and use information. AI does not have this limitation to the same extent.

“You have to ask yourself: ‘Okay, what is going wrong in some important sense with the man of systems?’ The man of systems has a very limited amount of compute–in some important sense. Right? This man of systems is a man. He can only understand systems that are so complex, he can have all the data in the world in front of him, he can’t actually meaningfully use that much of it… And, he is trying to be one man dictating all of these things. He’s got a hopeless task in front of him. He’s going to fail…However, when we’re talking about the AI, it doesn’t necessarily have to think about the bigger picture…”

The connecting vein between the dial of progress and the consistent failure of the man of system lies with individual separateness. There is no form of aggregate knowledge the man of system can access, because this only exists in the minds of distinct individuals. Because of this, progress flourishes when individuals, who have the most information on their specific strengths and preferences are left to pursue their vision for their own life. The man of system seeking to place his vision of progress on the rest of society fails to understand that individuals are ends in themselves, not means towards a unilaterally decided view of social welfare.

However, this does not mean that shifting the dial of progress towards requiring permission to innovate is not harmful. Though separateness ensures the man of system is doomed to fail in pursuing his own aims, by fixing the rules of the game, the aims of the individuals treated as chess pieces can be easily sacrificed at the altar of nationalism, corporate monopoly, or power acquisition. Although AI has the potential to push the dial upwards, a man of system could very well utilize the same technology to stifle further innovation and overall human freedom.

Related EconTalk Episodes:

Eliezer Yudkowsky on the Dangers of AI

Nick Bostrom on Superintelligence

Erik Hoel on the Threat to Humanity from AI

Dan Klein on The Theory of Moral Sentiments, Episode 6

Michael Munger on Permissionless Innovation

Related LF Network Content:

Katherine Mangu-Ward on AI: Reality, Concerns, and Optimism, Great Antidote Podcast

Harari and the Danger of Artificial Intelligence, by Pierre Lemieux, at Econlib

Accelerate Rather than Regulate Artificial Intelligence, by John O. McGinnis at Law and Liberty

Neoliberalism on Trial: Artificial Intelligence and Existential risk, by Walter Donway, at Econlib

ChatGPT and Economic Planning, by Pierre Lemieux, at Econlib