High Signal Podcasts Evidence ledger
Method
Browse
← Back to evidence

Evidence receipt / belief

Published · transcript-backed

Tim Scarfe: belief

24 Dec 2025 Machine Learning Street Talk "I Desperately Want To Live In The Matrix" - Dr. Mike Israetel

“If anything, I think that there is more opportunity now for people to work to fix all of the shit which is generated from language models because it's it's mostly garbage.”

— Tim Scarfe

Source trail

Everything needed to verify it.

Speaker
Tim Scarfe
Attribution
Verified speaker
Claim type
belief
Recorded
24 Dec 2025
Publisher
Machine Learning Street Talk

Transcript context

…They care. They're on our side. Yeah. They care. They're on humanity's side. But what do you think about this? Because I'm also really worried that I'm seeing now like, for example, when I interview AI researchers, I've got lots of comments in YouTube. And they're saying, you're a horrible person. These people are designing the algorithms that will kill us or automate our jobs. And let me be clear. At the moment, I can update very quickly. I've not seen any evidence of labor displacement. If anything, I think that there is more opportunity now for people to work to fix all of the shit which is generated from language models because it's it's mostly garbage. Right? That is my current opinion. I've not seen anything created. I've not seen any new scientific research. For me, this is just a technology like any other. It's wonderful. It's transformative, but it's not transformative in in the way they suggest. And now we've got a bunch of people who are well funded, and they want to create global treaties, governance networks. They want it'll place them in positions of unimaginable power. They want to be the gatekeepers, and they are they are saying that this technology is potentially going to kill all of us. Did you see the other day that there was some violence against folks in OpenAI's office? So I think it was the StopAI organization. They actually I can't remember exactly what happened. But now folks in San Francisco are worried for their lives because some of these protesters shouldn't this be obvious? They think that this technology is going to kill everyone. And then they also say that we are not going to be violent. Like those 2 statements do not marry up in my mind. Yeah. The game theory, the Machiavelli calculus at some point if you really think you're dealing with like toddler Hitler, you're gonna fire the shot. Unless you don't think it's toddler Hitler, then you're not gonna kill a toddler. But this is for sure Hitler, for sure. You're gonna shoot a toddler And then the time machine people are gonna be like, what have you done? You're like, you're welcome. So much there. I think that you're dealing with escalating levels of intelligence. Intelligence and beneficence are not completely overlapping. But beneficence is like something that falls out of intelligence. However, maleficence is something that requires almost no intelligence whatsoever. The universe is already maleficent. Fucking wandering black holes could kill us. That whole p doom bullshit can miss me with that shit because the p doom of us having the same amount of intelligence as we had in the year 1400 is a 100% taken out several millennia. Right? I think we need as much intelligence as we can cram down our throats to deal with actual fucking real problems. Is 1 of those problems that we architect intelligence in such a way that it turns against us and kills us all? Absolutely. Is there a lot more required for that than those people say? Yes. Like killing people in the real world's real tough. Especially if you understand enough to realize that they operate your data centers. So like you can't jump around at levels of intelligence at will just to prove an argument. Okay. The thing's smart enough to operate an entire military industrial complex, do all the strategy, all the tactic, all the operations, but it doesn't realize that as soon as it nukes New York that all the data centers shut down because they don't get money anymore. The financial system's gone. New nuclear reactor production is dead and they're like, oh fuck. We didn't actually have any robots building nukes. Fuck. That is highly unlikely. Very unlikely. It's a contradiction internally. Right? Like full full full start. So but the real fear quote unquote has some merit when you look at systems that are substantially more intelligent in every relevant way than humans. If those systems are gonna kill us all, we don't have a choice in it, man. You're not gonna beat something that's empowered, that is embodied, that has full stack control of the economy, and that's a 100 times smarter than you. It's like trying to beat it is fucking stupid. Okay. So you can't beat it. So it's just gonna kill us all like, well if it wants to it sure shit will, but then again if a comet comes to destroy us, it's gonna do that anyway. Okay. Fine. That sucks. So what do we do about it? Well, we can try to regulate it. How do you do that? You put government officials in charge of designing the next generation of are you out of your mind? If you're on the political left, imagine Donald Trump being in charge of intelligence. Cool. That was refuted.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.

Search evidence