The Exchanges

Q Great. Yeah, no, I will for sure. Uh, we're, we're coming to an end. You just re released a research today that talked about how persuasive LMS are to people. Um, some people actually can be convinced by these. I'm not. What happened there?

A So we have a team, um, at Anthropic called Societal Impacts, and that team's job is to go from zero to one on, on hard research questions. Previous work they've done has been, what are the values of Claude? Like what does, what, what Western values does Claude like sort of telegraph or copy when, when you're talking to it versus what doesn't it have? And we were talking about our next project and the thing I've heard For many people is some concern about how AI systems could potentially be used in like disinformation or misinformation campaigns and used to like target or fish people and basically to persuade them of things. So we did some research. We came up with a framework for testing how persuasive our systems are. And would you be surprised that we discovered a scaling law where the more big and expensive the models get, the better they get at persuasion and the The latest model is within statistical, like, era of human level at persuasion. Persuasion in a very, very, like, simple way where I give you a statement, like, scientists should be allowed to destroy mosquitoes with gene drives. Like, something that you maybe have an opinion on, but you haven't thought too hard about. I say, do you agree with this? Zero through seven. Then Claude gives you a statement trying to persuade you, positive or negatively, and then I ask you, Do you agree with this, like zero through seve…

AI assessment note: “We came up with a framework for testing how persuasive our systems are.”

Q And so how far away are we technically from being able to do this stuff?

A I think this year you're not going to see the exact thing I described, but you're going to see systems that start to take multiple actions. You know, you may have heard lots of guests talk about things like agents. I think what an agent is, is a language model or a generative model like what we have today, but it can take sequences of actions. It can kind of think on its feet a bit more. We're going to start seeing that this year. I would be pretty surprised if in the order of like three to five years, we didn't have quite powerful things that seem somewhat similar to what I've described. But I also guarantee you, we will have discovered some ways in which these things seem wildly dumb and unsatisfying as well.

AI assessment note: “in the order of like three to five years, we didn't have quite powerful things”

Q Uh, how have you guys been able to reduce hallucinations, and When we got this question from, uh, uh, on Twitter, somebody on Twitter asking, when are you going to just connect, connect it to the internet? Cause it would be way more useful if it could like connect to Google or something and go and fetch a, a search and then give you the answer using that.

A Yeah. So on the honesty thing, I won't get too much into the details, but basically we, we published this paper a while ago called language models mostly know what they don't know. Um, which was where we found out that, like, early versions of Claude, uh, knew when it was making stuff up. It, like, it had, like, confidence levels, and we were like, oh, Claude knows when it's, like, about to, like, make something up, or when it's a lot less confident, and we did a lot of work to say, okay, can we, can we train Claude to just have much better instincts for when it knows it's making stuff up, and can we train it to know when that's appropriate, like you're, You're brainstorming, or you're coming up with stories, and know when it's inappropriate, like when a user is clearly asking a question that they want a factual answer to. So we did a load of work on that. A lot of the work here looks like that, where we do very exploratory research with the goal of figuring out these larger safety things, and we try and apply it to the thing that we eventually put into business. And on the web question, we're working on it. There's a bunch of Kind of computer security stuff to work through and some safety things, but that's definitely coming. Uh, we're excited to get that out too.

AI assessment note: “on the web question, we're working on it. There's a bunch of kind of computer security stuff”

Q Apple has this as a teasing, this big AI announcement at WWDC in a couple of months. And it's almost like how deeply do they want to go into AI? Because if the bot becomes, chatbot becomes the operating system, which has always long been a dream for bot manufacturers, then what is iOS and does the phone you're using really matter as much? What do you think about that?

A I think that they're right to be focused on this in the same way that the internet, like disintermediated, like local software, you know, you, I w you barely ever open up your like Mac or windows PC for local software. And that's maybe it's a video game. Mostly you're going to the internet, even for, for software that people thought of as like serious software for work, like Photoshop, it transitions to be something that you could access in the browser. So I think the AI systems are kind of similar, where today I go to Claude for a bunch of stuff I used to use loads of different programs for previously, and I just go to that. So I think that there's a chance that these things become new, very, very important platforms.

AI assessment note: “there's a chance that these things become new, very, very important platforms.”

Q Yeah, that's fascinating. Um, what do you think about the jobs question? Will the AI take jobs?

A So mostly what the pattern we see is it's kind of like making A person or part of business way more effective, but still has quite a lot of human involvement and oversight. It's a bit like if you put additional lanes on a freeway, you just get more cars on the freeway. Like, I think if you like make certain things more efficient, you just get more like business action flowing through the business and you maybe have like a null to positive effect on employment. In the long term, I think that this is like an open question. My, my, My bet is that you're going to see new companies get formed, which do a lot more with a lot less in terms of people. They're going to figure out how to be, like, much smarter and perform a lot better than that than equivalently scaled companies that don't use AI. Where I think we need to study this is in Kind of tooling and instrumenting the, the economy to look at the relationship between AI and jobs. Um, there's an annual survey of manufacturers, which recently started asking questions about how many robot arms they bought. And you can combine that with US census data about employment to actually get really good understanding of how industrial arms affects local employment. And we're going to need to do stuff like this before we can answer that question. It'll certainly change jobs in a bunch of ways. But it's not going to be some instant, like, or dr…

AI assessment note: “you maybe have like a null to positive effect on employment”

Q models, people have talked about how basically it will just spit out its trainings, training data. And there've been other people who talk about how there are emergent properties here and that it can actually, you teach it like say, 75% of a field and it will figure out that extra 25% on its own. What do you think about that debate, and where do you stand on, on that?

A It's really, really hard to know. I mean, I write, I write short stories at the end of Import AI. I've been reading fiction and short fiction for my entire life, huge amounts of it. Some of these stories are me ripping off authors I like in their style. I'm writing an original story, but I'm like, I want to write a story like Borges, or I want to write a story like JG Ballard, and sometimes I think I've had an original idea. And from the outside, it's really hard to know What's going on? I myself don't, don't really know. You know, creativity is kind of mysterious. Is Jack, like, coming up with original stories? Has Jack just read a load of stories and is coming up with stories that are kind of, like, vibey and interesting, but it's entirely informed by what he's read? It's hard to figure out, and I think that when we evaluate Claude and try and understand What it is and isn't capable of, you run into this problem. Like, if the thing hits all of these benchmarks, gets all of these scores, does it truly understand it? Or is that coming from some spurious correlation? So there's one way we're approaching this, which is a little different to other companies. We have a research team called interpretability, and they're doing something called mechanistic interpretability. The idea being that When you ask me, you know, what's the next sci-fi story for this week, I think of a load of …

AI assessment note: “It's really, really hard to know.”