AI Doomers Are Driving a War on Computation | Adam Thierer artwork

AI Doomers Are Driving a War on Computation | Adam Thierer

MTS

October 14, 2026

Adam Thierer, Resident Senior Fellow at the R Street Institute, critiques effective altruism's approach to AI safety. Turn ideas into software people love. MTS is a live news and interview show covering technology, business, politics, and culture as it happens. Interested in sponsoring the show?

0:00

SPEAKER_1

And we are back, we are live with Adam Thierer, who's a resident senior fellow at the R Street Institute, a visiting senior fellow at FIRE, the author of Permissionless Innovation. Very topical at the moment, Adam, welcome to MTS.

0:14

Adam Thierer

Thanks for having me back.

0:16

SPEAKER_1

So tell us about your recent article on effective altruism, thoughts on this century's biggest idea. I think you have some of the best criticism of EA that I have seen, because it is thoughtful, and it takes the ideas behind it seriously, and it's not like just total shutting it down. But it's very good. So tell the audience more about it.

0:38

Adam Thierer

Yeah. Well, thanks for saying that. Well, first of all, effective altruism is rooted in utilitarianism, and there's a lot we can learn from utilitarianism and a lot of benefits from that way of thinking. But the problem is when one seeks to convert utilitarianism into a quasi-religion, and treat morality like a math problem to be solved through various types of interventions and elitist planning. Well, that's a recipe for real trouble for society and for economies.

Too many people in the EA community, in my opinion, suffer from a sort of fatal conceit of thinking that really complex moral, social, and technological problems can be solved through rational optimization and enlightened central planning. We've seen this script before. It's very well-intentioned.

But again, morality is not a math problem in society and economies can't be easily optimized. It's more complicated than that.

1:34

SPEAKER_1

Yeah. Do you think this is like, with alignment, I think different people have different ideas as to what alignment actually means. Some people literally do picture it as a math problem to be solved. Others see it as more of like a defense in depth type approach where like, in some sense, there is some math problem inevitably that results from like applying morality to machine learning. But that's not like the only technique that we have in order to make systems behave how we want and not do bad things.

There are other things in different layers like cybersecurity that we'll need to do as well.

2:13

Adam Thierer

Yeah, that's right. I mean, solving for safety in complex systems has always been really, really difficult and it cannot always easily be deduced to a hard science or concrete math. And unfortunately, I think in the EA world, we see a lot of speculative catastrophism around the idea of AI doom or what they call their P Dune numbers, right? Everybody wants to know what's your P Dune? What's your P Dune? And I'm never one to actually play that game because I think we're talking about subjective probabilities or feelings dressed up as hard math or numbers. And the reality is that speculative futures are just that, just because there's a percentage after these guesses does not mean that they're hard science. Everything's a lot more complicated. It doesn't mean we shouldn't think about those risks. We should absolutely think about those risks, but we shouldn't sell it to the public and policy makers like, we have exactly a 10 percent chance of all dying within the next five to 10 years.

That's silliness, that's fearing, that's fueling the panic that we see around AI and robotics and advanced computation right now. Again, that's sort of like what I call speculative catastrophism, masquerading science, and that's bad for us.

3:22

SPEAKER_1

So what do you think would be more productive ways for people to go about raising public salience around and addressing AI risks that do exist?

3:31

Adam Thierer

Well, the good thing that the EA movement has been teaching us is that we need to take these AI safety risks seriously and we need to come up with a good playbook for addressing some of them and I can agree with a lot of their prescriptions and ideas and some of them are already in place. We already see a lot of really good efforts being made. The whole process of reinforcement learning involves essentially a sort of alignment process on an ongoing iterative way. And a big point of the piece that I wrote on my critique of Effective Altruism is that safety is not a static concept. It's not a one and done thing. Real safety comes from ongoing experimentation and iterative trial and error, not through one shot centralized controls and interventions.

It takes a more decentralized bottom up approach to get safety right. It's very messy. People want simple playbooks. They want silver bullets, especially policy makers like tell us the plan. What should we do? It's not always that easy, especially when it comes to complex matters like cyber security of advanced digital and computational system. That's where it gets incredibly difficult.

Read the other 6 minutes of this episode here.

$15 gets you this episode and 99 more. Credits never expire. After checkout, one link brings you back here to read it.

Not what you expected? Full refund within 14 days.

Thousands of transcripts fetched by people building searchable podcast archives

Fetch transcripts with code

The same credits work from a script or an AI agent. The demo key returns a sample episode in full, no card needed:

request
curl -H "x-api-key: pt_demo" \
  https://spoken.md/transcripts/1000651996090

Markdown with the speakers named, for your notes, your knowledge base, or anything that makes HTTP calls.

From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.

Using your own key:

request
curl -H "x-api-key: YOUR_KEY" \
  https://spoken.md/transcripts/1000796329266