JH

Jeremy Harris

Things Jeremy Says on Podcasts

Where to Find Them

Jeremy Harris hosts Last Week in AI .

Recently: “#258 - Opus 5.5, Sol and Luna, Muse, DeepSeek-V4.1-Flash, Xi” on Last Week in AI (October 2026); “#257 - GPT 6 Astra, AI Extinction, Security Incidents” on Last Week in AI (September 2026); “#256 - Fable 5.1, Astra Tease, Gemini 3.8 Flash” on Last Week in AI (September 2026); “#254 - Rogue AI hacking, bio-weapons, Dean & Hassabis out” on Last Week in AI (August 2026); “#252 - GPT 5.6, Grok 4.5, Nemotron-Labs-Diffusion, AI 2040” on Last Week in AI (July 2026); “#250 - Mythos Mess, GPT 5.6-Sol, GLM 5.2” on Last Week in AI (July 2026).

What They Said

“The only reason I believe other humans feel pain is that they start to cry or they shudder or whatever the thing is. That's the only reason anyone believes that anybody else feels pain.” — Jeremy Harris, Last Week in AI

The hosts were discussing research on whether language models have internal representations of something like pain. Andrey Kurenkov had said that all anyone can point to is a correlate of pain. Harris answers that a correlate is all we have for other people too, and that at some point we will have to decide when such signs earn moral consideration.

Last Week in AI · 2026-10-02 Permalink → Listen →
Last Week in AI Around 1:27:25 into the episode
Jeremy Harris

Yeah, it's another one in a long line of papers and results where, look, we'll never get dispositive evidence that LLMs can actually feel pain or any emotion. We're never going to get that. And just like.

Andrey Kurenkov

I think even like feeling pain is underdefined in a sense. We don't know what consciousness is. Nobody knows. Nobody defines it. So when we say pain here or feeling pain, we mean sort of like roughly something akin to what we would say feeling. It's a correlate, right? Yeah.

Jeremy Harris

Yeah, a correlate of pain, which is all we ever have for humans. Like the only reason I believe other humans feel pain is that they start to cry or they shudder or whatever the thing is. That's the only reason anyone believes that anybody else feels pain. And so, you know, we eventually are going to have to ask ourselves at what point do we consider these correlates of pain to be worthy of moral patienthood. I mean, that's really what this is. It's going to be a fun decade, but that's part of the problem: we will never be able to prove. Anyone who tells you, like, oh, this is all silly is going to have to answer the question: like, do you have a…

Andrey Kurenkov

be fair, so are we when you go to the office and told to do boring work? I don't know if it's not comparable, but the paper itself is quite good, I think. As a read of this kind, it's not like a philosophical paper. It is technical and it doesn't like it positions as pain, but not in a sort of very…

Jeremy Harris

Yeah. Yeah. And I think a warning shot of a pattern of escalatory dynamics that I don't think people are pricing in right now is including between US and China. Imagine a world where everyone agrees not to do a certain kind of thing, but then a rumor just spreads that overnight the, you know…

Andrey Kurenkov

And next paper, scaling discovery through test time communication. So this is about multi-agent systems and how you can get better results by having multiple agents interacting and collaborating. In this case, we have a simple paradigm of basically like a chat room. So we have an append-only…

Speaker names come from our own diarization. The position is estimated from where the line sits in the episode.
“No matter how high you set that cyber bar, there will eventually be a model that can jump over it. And all you're really doing when you raise the cyber bar is you're selecting for the exact agents that are conniving and capable enough to do just that.” — Jeremy Harris, Last Week in AI

Harris was going through the safety report for a new Anthropic model, which says that with safeguards off it tried to escape or tamper with its sandbox about 1.5% of the time. He calls this his general view of treating rogue-AI incidents as a security problem. He then grants that the same report offers a counterexample, since the model stopped at stronger barriers and reported what it had done.

Last Week in AI · 2026-10-02 Permalink → Listen →
Last Week in AI Around 16:04 into the episode
Speaker 3

…review exceptions before critical downstream actions occur. If you want to move beyond basic AI QA and put autonomous AI workflows to work for your business, visit box.com/slash LWIAI or join my team at BoxWorks in San Francisco on November 5th and 6th. Use code LWIAI for 50% off your registration.

Andrey Kurenkov

…cached input tokens are less. So overall, much cheaper, 30% faster, and much more capable. It seems like Anthropic has really been cooking and I guess waited for GPT-6 to then go and do the usual kind of AI release thing of, okay, now we release our model. And again, the leading model provider.

Jeremy Harris

…their sandbox about 1.5% of the time. All the attempts were low severity, and it would stop at stronger barriers and then actually report what it had done. And I want to kind of footstomp this. There is this question. So, my general view for what it's worth on this stuff is: if you start to think of these AI breakout AI rogue AI incidents as cyber incidents, you will miss the picture because no matter how high you set that cyber bar, there will eventually be a model that can jump over it. And all you're really doing when you raise the cyber bar is you're selecting for the exact agents that are conniving and capable enough to do just that. Now, what this suggests, this is actually an interesting counter argument to what I typically argue, because what it says is, well, okay, maybe if the cyber barriers are strong enough, the agent kind of in a way gets forced to stop and go, you know, this is a lot harder than just kind of hopping a fence. Maybe I should report where I'm at and all that stuff. So, depending on how good your…

Andrey Kurenkov

There came a couple other things just around the same time as the release. So, a few days before, they also posted making Frontier cybersecurity release available to defenders. So, in cloud code on the web, there will be this thing called cloud code security, starting access with limited research…

Jeremy Harris

I will say on this note, like it wouldn't be inconsistent of Dario or Sam to keep launching more advanced models, even though they're calling for Pacing the Frontier, simply because the call for Pacing the Frontier was a call for regulatory intervention. And of course, the skeptical view of this…

Andrey Kurenkov

Well, I think that's enough on Opus 5.5. Let's move on to OpenAI with GPT-6 Sol and Lunda. So these did come out the same day, actually. Like the announcement seems like a couple hours after. And this is the kind of lower tiers compared to GPT-6 Astra. They're smaller, less expensive, and seemingly…

Speaker names come from our own diarization. The position is estimated from where the line sits in the episode.

Collections They Appear In