Synthetic Socrates?

Synthetic Socrates?

-Allen Stairs

Jimmy Licon completed his PhD in the Philosophy Department in 2019 and is an Assistant  Teaching Professor at Arizona State University. Jimmy is a good-natured contrarian, exploring what he calls "uncommon wisdom." He's finishing a book making a case that sometimes it's better to know less than more. He thinks that even if you don't believe there's a God, it might be rational to hope there is. And he argues that whether or not wearing masks to prevent the spread of Covid is a good thing, a widely-repeated argument that we're obliged to proves too much. One goal of Jimmy's work is to take a careful look at our routine assumptions. 

I caught up with Jimmy after seeing a Facebook post with a link to a paper he wrote (forthcoming in Think | Cambridge Core) called "Synthetic Socrates and the Philosophers of the Future."  The idea is that some day, maybe not so far away, we might be able to offload philosophical research onto AI devices. Here are some excerpts from our conversation.

Me: What made you decide to think about Synthetic Socrates?

Jimmy: It's actually an idea that occurred to me while I was at Maryland. I thought: if brains are computational, what's stopping us from building an AI philosopher? If you think philosophy is about accumulating truths, as opposed to being like therapy, then you might think there's more or less impetus to offload our work to AI. 

Anyway, the grad students mocked me pretty roundly for this over the rest of my time at Maryland. [Chuckles] 

Me: Mocked you in the sense of being skeptical of the idea?

Jimmy: I think it was a mixture of skepticism and "If you believe this, what are you here for?" 

Me: When you talk about offloading to a computer, is the interest specifically in whether a computational device could do philosophy, or just some sort of physical thing of whatever sort? 

Jimmy: I find the possibility of AI more credible than the possibility of aliens landing on earth and being smarter than we are and more capable than we are but we somehow convince them to do our philosophy for us. 

Me: Yes. In some sense there's a more serious question about computers than about aliens or whatnot. But here's another question. If we built an artificial physicist, we have criteria for deciding whether the artificial scientist comes up with something true. But in the case of the artificial philosopher, can we imagine getting to a point where we think: "Well I don't know how it got there, but I have to believe that somehow it got the answer right." What would it mean for the AI to "solve the free will problem?"

Jimmy: With physics or medicine, we'd have benchmarks and empirical criteria. But if the question you're asking proved too difficult to answer, then that might put pressure on the metaphilosophical view that what we're doing is accumulating truths. Maybe philosophy is more therapeutic or more related to epistemic humility than to coming to a particular conclusion. Maybe philosophy is a disposition that you follow -- a way of thinking about the world. 

On the other hand it could be that we evaluate what the AI philosopher says in the way we evaluate what other human philosophers say. And it's also possible that if our benchmarks for evaluating philosophy weakened with AI philosophers, it might be because they're introducing a kind of intellectual diversity. The AI might give us a kind of new input, so to speak. Or it might be that the AI is a very fruitful, helpful interlocutor and building them will make the human philosophers that much better. For example: in chess, human teams and computer teams sometimes lose to hybrid teams.

Me: On whether Synthetic Socrates gets things right: for at least some philosophical problems -- like free will -- I'm skeptical of the idea that there's such a thing as the correct answer. But what do you think?

Jimmy: What someone thinks of Synthetic Socrates will depend on what they think philosophy is for. I know why I do philosophy. I do it because it's fun. I don't care if Synthetic Socrates does it better than I do. I'm not trying to accumulate a bunch of truths. I do it because I can't help but do it!

-----

For more of what Jimmy is working on, try his Substack: 

https://jimmyalfonsolicon.substack.com/