If AI is truly very powerful, is the chance that it will save us from other existential risks higher than the chance it will kill us?

The question above posed by polymath Tyler Cowen in this episode, is the core of his conversation with host Russ Roberts. Roberts begins by asking Cowen what he thinks the biggest impacts Artificial Intelligence will have on the economy and elsewhere. The conversation covers the possibility of intelligence and/or wisdom in AI, and its potential “cognitive oomph.” (That’s a technical term, of course.)

It’s always a treat to sit in on a conversation between these two. And once you’ve done so, we hope you’ll consider the prompts below. As you know, we ♥ to hear from you!



1- Cowen suggests that the closest historical analogs to AI are the printing press and electricity, stressing that major technological advances tend to be disruptive. But with great disruption, they often bring highly significant benefits. The question is, according to Cowen, how do you face up to them?

Are we willing to tolerate major disruptions which have benefits much higher than costs, but the costs can be fairly high? What other historical analogs can you suggest that help shed light on the trade-offs we may anticipate with widespread adoption of AI? What sort of disruptions do you think AI will bring to bear, and how much should be tolerated, and why?


2- Roberts asks Cowen where he sees the greatest impact of AI now. Cowen suggests it can already serve as an “incredible interactive tutor,” and that it will soon eliminate a great deal of back office work. Roberts and Cowen agree that the impact on coding is already significant, and then Roberts turns to poetry.

Where do you see the greatest impact short-term effects of AI? To what extent do you agree with the suggestions offered by Cowen and Roberts? Would you let ChatGPT write a condolence note? What about a love poem? Why or why not?


3- Cowen asserts that alongside AI, how you behave in real life will become even more important. So back to poetry. Recall the story of Cyrano de Bergerac. How does this illustrate the need for transparency, and what does this suggest about real-life versus virtual engagement in things like dating apps? The classroom? Employment situations?


4- Roberts asks Cowen if there anything he would regulate or  slow down with regard to AI. Cowen continually insists that it’s not whether to regulate, but who should regulate and how. It’s about decentralization, he says- ensuring proper checks and balances and mobilizing decentralized knowledge.

What does Cowen mean when he says that Hayek and Polanyi should be at the center of the debate, rather than Arrow and Bentham? What role does he see for social norms in constraining AI? What role do you see for regulation and/or social norms? If there’s such an emphasis on decentralization, why is Cowen so adamant that we need to employ scientific modeling in the conversation?


5- Unlike several other recent guests, Cowen is much more optimistic about the future of AI, but not that it doesn’t come without risk. He says, “There are these very serious risks, but there’s also a Millenarian tendency in human thought that tends to rise in volatile times. And, we have to think about how to deal with those Millenarian tendencies. They are themselves a form of risk: that our reaction to an event can be worse than the event itself, even if the event involves some very high costs.” What does this mean? Which do you think is more dangerous- AI itself, or the doomsday predictions it inspires? Explain.


Bonus Question: Says Cowen, “One of my predictions is that GPT models will raise the relative–and indeed–absolute wages of carpenters.” Why might this be the case? Double Bonus if you use supply and demand graphs in your explanation!