Tech Brew Ride Home - (BNS) Simon Willison And SWYX Tell Us Where AI Is In 2025

Episode Date: January 11, 2025

The great Simon Willison joins SWYX and I to talk about everything we learned about LLMs in 2024, and what the state of AI is generally, as we go into 2025. Here is Simon's blog post we keep referring to: https://simonwillison.net/2024/Dec/31... 00:00 The State of AI in 2025 10:05 The Evolution of AI Models 19:54 Challenges in AI Agents 30:07 The Future of AI in Creative Industries 38:29 The Rise of AI Influencers 40:54 Credibility in the Age of AI 43:15 The Future of User Interfaces for LLMs 51:17 Local LLMs and Desktop AI Applications 55:17 AI Tools and Applications for Everyday Use 01:01:26 The Future of OpenAI and AI Regulation 01:08:08 The Need for Better Criticism of LLMs 01:10:41 The Future of Wearables and AI Integration Learn more about your ad choices. Visit megaphone.fm/adchoices

On April 4th, 2023, around 2 in the morning, a man was found stabbed multiple times on a sidewalk in downtown San Francisco. Hey, who did this to you? What happened next turned the story into a political firestorm. Reports have identified the victim as Bob Lee, the founder of Cash App. From Bloomberg Podcasts, this is Foundering, the Killing of Bob Lee, beginning April 16. Welcome to the first bonus episode of the TechMeme Right Home for the year 2025. I'm your host, as always, Brian McCullough. Listeners to the pod over the last year know that I have made a habit of quoting from Simon Willison when new stuff happens in AI from his blog.

Simon has become a go-to for many folks in terms of, you know, analyzing things, criticizing things, criticizing things. in the AI space. I've wanted to talk to you for a long time, Simon, so thank you for coming on the show. No, it's a privilege to be here. And the person that made this connection happen is our friend Swix, who has been on the show back, even going back to the Twitter space's days, but also an AI guru in their own right. Swix, thanks for coming on the show also. Thanks. I'm happy to be on and have been a regular listener. just happy to contribute as well. And a good friend of the pod, as they say.

All right, let's go right into it. Simon, I'm going to do the most unfair, broad question first, so let's get it out of the way. The year 2025, broadly, what is the state of AI as we begin this year? Whatever you want to say, I don't want to lead the witness. Wow, so many things, right? I mean, the big thing is everything's got really good and fast and cheap. Like that was the trend throughout all of 2024. The good models got so much cheaper.

They got so much faster. They got multimodal, right? The image stuff isn't even a surprise anymore. They're growing video, all of that kind of stuff. So that's all really exciting at the same time. They didn't get massively better than GPT4, which was a bit of a surprise. So that's sort of one of the open questions is, are we going to see? But I kind of feel like.

that's a bit of a distraction because GPT4, but way cheaper, much larger context lengths, and it can do multimodal, is better, right? That's a better model, even if it's not, yeah. What people were expecting or hoping, maybe not expecting, not the right word, but hoping that we would see another step change, right? Right. From like GPT2 to three to four, we were expecting or hoping that maybe we were going to see the next evolution in that sort of, yeah.

We did see that, but not in the way we expected. We thought the model was just going to get smarter. And instead, we got massive drops in price. We got all of these new capabilities. You can talk to the things now, right? They can do simulated audio input, all of that kind of stuff. And so it's kind of, it's interesting to me that the models improved in all of these ways we weren't necessarily expecting. I didn't know it would be able to do an impersonation of Santa Claus.

And I could talk to it through my phone and show it what I was seeing by the end of 2024. But yeah, we didn't get that GPT5 step. And that's one of the big open question. is, is that actually just around the corner and we'll have a bunch of GPT5 class models drop in the next few months? Or is there a limit? If you were a betting man and wanted to put money on it, do you expect to see a phase change, step change in 2025? I don't particularly for that, like, the models, but smarter. I think all of the trends we're seeing right now are going to keep on going, especially the inference time compute, right, the trick that 01 and 03 are doing, which means that

you can solve harder problems, but it costs more and it churns away for longer. I think that's going to happen because that's already proven to work. I don't know. I don't know. Maybe there will be a step change to a GPT5 level. But honestly, I'd be completely happy if we got what we've got right now, but cheaper and faster and more capabilities and longer contexts and so forth. That would be thrilling to me.

Digging in to what you've just said, one of the things that, by the way, I hope to link in the show notes to Simon's. year-end post about what things we learned about LLMs in 2024. Look for that in the show notes. One of the things that you did say that you alluded to even right there was that in the last year you felt like the GPT4 barrier was broken like i.e. other models, even open source ones are now regularly matching sort of the state of the art. Well, it's interesting right. So the GPT4 barrier was a year ago. The best available model was opening as GPT4 and nobody else had even come close to it. And they'd been in the lead for like nine months, right? That thing came out

in what, February, March of 2023. And for the rest of 2023, nobody else came close. And so at the start of last year, like a year ago, the big question was, why has nobody beaten them yet? Like, what did they know that the rest of the industry doesn't know? And today, that I've counted 18 organisations other than GPT4, who've put out a model which clearly beats that GPT4 from a year ago thing. Like maybe they're not better than GPT40, but that's, that, that barrier got completely smashed. And yeah, a few of those I've run on my laptop, which is wild to me. Like, it was very clear, it felt very clear to me a year ago that if you want GP4, you need a rack of $40,000 GPUs just to run the thing. And that turned out not to be true. Like the, this is that big trend from last year of the models

getting more efficient, cheaper to run, just as capable with smaller weights and so forth. I ran another GPT4 model on my laptop this morning, right? Microsoft 5-4 just came out. And that, if you look at the benchmarks, it's definitely, it's up there with GPT-40. It's probably not as good when you actually get into the vibes of the thing. But it runs on my, it's a 14-gigabyte download and I can run it on a MacBook Pro. Like, who saw that coming? The most exciting, like the close of the year on Christmas Day, just a few weeks ago, was when Deepseek dropped their Deepseek V3 model on Huggy.

face without even a read-me file. It was just like a giant binary blob. That I can't run on my laptop. It's too big. But in all of the benchmarks, it's now by far the best available open weights model. Like it's beating the meta-lamas and so forth. And that was trained for $5.5 million, which is a tenth of the price that people thought it cost to train these things. So everything's trending smaller and faster and more efficient. Well, okay, I kind of was going to get to that later. but let's combine this with what I was going to ask you next, which is, you know, you're talking also in the piece about the LLM prices crashing, which I've even seen in projects that I'm working on, but explain that to a general audience because we hear all the time

that LLMs are eye-wateringly expensive to run. But what we're suggesting, and we'll come back to the cheap Chinese LLM, but first of all, for the end user, what you're suggesting is that we're starting to see the cost come down sort of in the traditional technology way of cost coming down over time? Yes, but very aggressively. I mean, my favorite thing, the example here is if you look at GPT3, so OpenAI's GPT3, which was the best developed model in 2002 and through most of 2023, the models that we have today, the OpenAI models, are 100 times cheaper.

So there was a 100x drop in price for OpenAI from their best available model like two and a half years ago to today. Just to be clear, not to train the model, but for the use of tokens and things. Exactly, for running prompts through them. And then when you look at the really, the top tier model providers right now, I think are OpenAI, Anthropic, Google and Meta. And there are a bunch of others that I could list there as well. Mistral are very good. The deep-seeking quen models have got great.

There's a whole bunch of providers serving really good models. But even if you just look at the sort of big brand name providers, they all offer, models now that are a fraction of the price of the models we were using last year. I think I've got some numbers that I threw into my blog entry here. Yeah. Like Gemini 1.5 Flash, that's Google's fast, high quality model, is how much is that? It's 0.075 dollars per million tokens. Like these numbers are getting so difficult to compare.

Sense per million now. Sense per million. Sense per million makes a lot more sense. They have one model, 1.5 Flash 8B, the absolute cheapest of the Google models, is 27 times cheaper than GPT 3.5 Turbo was a year ago. Like, that's it. And GPT 3.5 Turbo, that was the cheap model. Right now we've got something 27 times cheaper. And this Google one can do image recognition, it can do million token context, all of those tricks. It's very, it really is startling how inexpensive some of this stuff has got. Now, are we assuming that this, that happening is directly the result of competition? Because again, you know, Open AI and probably they're doing this for their own almost political reasons, strategic reasons, keeps saying, we're losing money on everything, even the $200. So they probably wouldn't, the prices wouldn't be coming down if there wasn't intense competition in the space.

So the competition is absolutely part of it. But I have it on good authority from sources I trust that Google Gemini is not. operating at a loss, like the amount of electricity to run a prompt is less than they charge you. And the same thing for Amazon Nova, like somebody found an Amazon executive and got them to say, yeah, we're not losing money on this. I don't know about Anthropic and Open AI, but clearly that demonstrates it is possible to run these things at these ludicrously low prices and still not be running at a loss if you discount the army of PhDs and the training costs and all of that kind of stuff. One more for me before I let Swix jump in here. To come back to Deep Seek and this idea that you could train, you know, a cutting edge model for $6 million, I was saying on the show like six months ago that if we are getting to the point where each new model costs a billion, $10 billion, $100 billion to train that at some point, it would almost only nation states would be able to train the new models.

Do you expect what Deep Seek and maybe others are proving to sort of blow that up? or is there like some sort of a parallel track here that maybe I'm not technically, I don't have the now to understand the difference? Is the model, are the models going to go up to $100 billion, or can we get them down sort of like DeepSeek has proven? So I am the wrong person to answer that because I don't work in a lab training these models. So I can give you my completely uninformed opinion, which is I feel like the deep seek thing, that was a bombshell. That was an absolute bombshell. When they came out and said, hey, look, we've trained one of the best available models and it cost a six, five, and a half million dollars to do it.

I feel, and one of the reasons that's so efficient is that we put all of these export controls in to stop Chinese companies from buying GPUs. So they were forced to go as efficient as possible. And yet, the fact that they've demonstrated that that's possible, I think it does completely

tear apart this mental model we had before, that yeah, the training runs just keep on getting more and more expensive and the number of organizations that can afford to run these training runs keeps on shrinking. That's been blown out of the war. So yeah, that's, again, this was our Christmas gift. This was the thing they dropped on Christmas Day.

Yeah, it makes me really optimistic that we can, there are, it feels like there was so much low-hanging fruit in terms of the efficiency of both inference and training. And we spent a whole bunch of last year exploring that and getting results from it. I think there's probably a lot left. I think there's probably, I would not be surprised to see even better models trained spending even less money over the next six months. Yeah.

So I think there's an unspoken angle here on what exactly the Chinese labs are trying to do. Because deep sea made a lot of noise around the fact that they train their model for $6 million. And nobody quite believes them. Like it's very, very rare for a lab to trumpet the fact that they're doing it for so cheap. They're not trying to get anyone to buy them. So why are they doing this? They make it very, very obvious that their lab, you know, Deepseek is about 150 employees. It's an order of magnitude smaller than at least anthropic and maybe maybe more so for OpenEI.

And so what's the end game here? Are they just trying to show that the Chinese are better than us? So Deep Seek, it's the arm of a hedge. It's a quant fund, right? It's an algorithmic quant trading thing. So I would love to get more insight into how that organisation works. My assumption from what I've seen is it looks like they're basically just flexing. They're like, hey, look at how utterly brilliant we are with this amazing thing that we've done. And it's working, right? And so is that it? Is this just their kind of like, this is why our company is so amazing, look at this thing that we've done?

I don't know. I'd love to get some insight from within that industry as to how that. all playing out. The prevailing theory among the local Lama crew and the Twitter crew that I indexed for my newsletter is that there is some amount of copying going on. It's like Sam Altman, you know, tweeting about how they're being copied. And then also there's this, there are the other sort of opening eye employees that have said stuff that is similar that Deep Seeks rate of progress is how US intelligence estimates the number of foreign spies embedded in top labs. Because a lot of these ideas do spread around, but they surprisingly have a very high density of them in the

DeepSeek V3 technical report. So it's interesting. We don't know how much, how many, how much tokens. I think that, you know, people have run analysis on how often Deep Seek thinks it is clod or thinks it is opening a GPC4. And we don't, we don't know. We don't know. I think for me, like, yeah, we basically will never know as external commentators. I think what's interesting is how, where does this go. Is there a logical floor or bottom? By my estimation, for the same amount of ELO, started last year to the end of last year, cost went down by a thousand X for the GPT4 intelligence. Do they go down a thousand X this year? That's a fascinating question. Yeah. Is there a Moore's law going on, or did we just get a one-off benefit last year for some weird reason? My uninfirm

formed hunch is low-hanging fruit. I feel like up until a year ago, people haven't been focusing on efficiency at all. It was all about what can we get these weird shaped things to do. And now once we've sort of hit that, okay, we know that we can get them to do what GPT4 can do, when thousands of researchers around the world all focus on, okay, how do we make this more efficient? What are the most important? Like, how do we strip out all of the weights that have stuff and that doesn't really matter, all of that kind of thing? So yeah, maybe that was a, maybe 2024 was a freak year of all of the low-hanging fruit coming out at once. And we'll actually see a reduction in that rate of improvement in terms of efficiency.

I wonder. I mean, I think we'll know for sure at about three months' time if that trend's going to continue or not. Yeah, I agree. You know, I think the other thing that you mentioned the DCV3 was the gift that was given from Deepseek over Christmas. But I feel like the other thing that might be underrated was Deepseek R1. which is a reasoning model you can run on your laptop. And I think that's something that a lot of people are looking ahead to this year. Did they release the weights for that one?

Yeah. Oh, my goodness, I missed that. I've been playing with Quinn. So the other great, the other big Chinese AI Lab is on Alibaba's Quinn. Actually, yeah. I sorry. Yes. No, R1 is an API available.

Yeah. Exactly. When, that's really cool. So Alibaba's Quinn have released two reasoning models that I've run in my laptop now. The first one was key Q. WQ and then the second one was QVQ because the second one's a vision model. So you can like give it vision puzzles and a prompt that these things, they are so much fun to run because they think out loud. It's like the opening R.01 sort of hides its thinking process. The credit ones don't.

They just, they just churn away. And so you'll give it a problem and it will output literally dozens of paragraphs of text about how it's thinking. My favorite thing that happened with QWQ is, I asked it to draw me a pelican on a bicycle in SVG. That's like my standard stupid prompt. And for some reason, it thought in Chinese. It spat out a whole bunch of like Chinese text onto my terminal on my laptop. And then at the end it gave me quite a good sort of artistic pelican on a bicycle. And I ran it all through Google Translate. And yeah, it was like, it was contemplating the nature of SVG files as a starting point. And the fact that my laptop can think in Chinese now, is so delightful. It's so much fun watching it do that.

Yeah, I think Andre Carpathie was saying, you know, we know that we have achieved proper reasoning inside of these models when they stop thinking in English. And perhaps the best form of thought is in Chinese. But yeah, for listeners who don't know, Simon's blog, he always, whenever a new model comes out, I don't know how you do it, but you're always the first to run Pelican bench on these models. I just did it for 5'4 this morning. Yeah. So I really appreciate that. You should check it out. These are not theoretical. Like Simon's blog actually shows them.

Let me put on the investor hat for a second because from the investor side of things, a lot of the VCs that I know are really hot on agents. And this is the year of agents. But last year was supposed to be the year of agents as well. lots of money flowing towards agentic startups. But in your piece that, again, we're hopefully going to have linked in the show notes, you sort of suggest there's a fundamental flaw in AI agents as they exist right now. Let me quote you, and then I'd love to dive into this.

You said, I remain skeptical as to their ability based, once again, on the challenge of gullability. LLMs believe anything you tell them, any systems that attempt to make meaningful decisions on your behalf will run into the same roadblock. how good is a travel agent or a digital assistant or even a research tool if it can't distinguish truth from fiction? So essentially what you're suggesting is that the state of the art now that allows agents is still, it's still that sort of 90% problem, the edge problem getting to the, or is there a deeper flow? What are you saying there? So this is the fundamental challenge here. And honestly, my frustration with agents is mainly around definitions. Like, if you ask anyone who says they're working on agents to define agents,

you will get a subtly different definition from each person. But everyone always assumes that their definition is the one true one that everyone else understands. So I feel like a lot of these agent conversations, people talking past each other, because one person's talking about the sort of travel agent idea of something that books things on your behalf. Somebody else is talking about LMs with tools running in a loop with a cron job somewhere. And all of these different things, you ask academics and they'll laugh at you because they've been debating what agents mean for over 30 years at this point. It's like this long-running, almost sort of an in-joke in that community. But if we assume that for this purpose of this conversation, an agent is something which

you can give a job and it goes off and it does that thing for you, like booking travel or things like that, the fundamental challenge is it's the reliability thing which comes from this gullibility problem. And a lot of my interest in this originally came from when I was thinking about prompt injection, a sort of this form of attack against LLM. systems where you deliberately lay traps out there for this LLM to stumble across. And which I should say you have been banging this drum that no one's gotten any far, at least on solving this that I'm aware of, right? That's still an open problem.

We've been talking about this problem. And like a great illustration of this was Claude. So Anthropic released Claude computer use a few months ago. Fantastic demo. You could fire up a Docker container and you could literally tell it to do something and watch it, open a web browser and navigate to a web browser. and navigate to a web page and click around and so forth. Really, really, really interesting and fun to play with.

And then one of the first demo somebody tried was, what if you give it a webpage that says, download and run this executable? And it did, and the executable was malware that added it to a botnet. So the very first, most obvious, dumb trick that you could play on this thing just worked, right? So that's obviously a really big problem. If I'm going to send something out to book travel on my behalf, I mean, it's hard enough for me to figure out which airlines are trying to scam me and which ones aren't. Do I really trust a language model that believes the literal truth of anything that's presented to it to go out and do those things?

Yeah, I definitely think there's, it's interesting to see Anthropic doing this, because they used to be the safety arm of opening eye that split out and said, you know, we're worried about, you know, letting them. this thing out in a while, then here they are in enabling computer use for agents. It feels like things have merged. I'm also fairly skeptical about, you know, this always being the year of Linux on the desktop. And this is the equivalent of this being the year of agents, that people are not predicting so much as wishfully thinking and hoping and praying for their companies and agents to work.

But I feel like things are coming along a little bit. to me it's kind of like self-driving. I remember in 2014 saying that self-driving was just around the corner. And I mean, it kind of is, you know, like in the Bay Area. And then you get in a Waymo and you're like, oh, this works. Yeah, but it's a slow cook. It's a slow cook. Over the next 10 years, we're going to hammer out these things.

And the cynical people can just point to all the flaws. But like, there are measurable or concrete progress steps that are being made by these builders. There is one form of agent that I believe in. I believe, mostly believe in the research. assistant form of agents. Yes, I was going to see. We've got a difficult problem. And I've got, like, I'm on the beta for the Google Gemini 1.5 Pro with deep research, I think

it's called. Oh, God. These names. These names, right? But I've been using that. It's good, right? You can give it a difficult problem and it tells you, okay, I've gone to look at 56 different websites, and it goes away and it dumps everything into its contacts, and it comes up with

a report for you. And it's not, it won't work against adverse. websites, right? If there were websites with deliberate lies in them, it might well get caught out. Most things don't have that as a problem. And so I've had some answers from that, which were genuinely really valuable to me. And that feels to me like, I can see how given existing LM tech, especially with Google Gemini with its like million token contacts, and Google with their crawl of the entire web and they've got like search, they've got search and they've got a cash of every page and so forth. That makes sense to me. And that what they've got,

right now. I don't think it's not as good as it can be, obviously, but it's a real useful thing, which they're going to start rolling out. So, you know, like perplexity, you've been building the same thing for a couple of years. That, that I believe in. You know, if you tell me that you're going to have an agent that's a research assistant agent, great. The coding agents, I mean, chat GPT code interpreter, nearly two years ago, that thing started writing Python code, executing the code, getting errors, rewriting it to fix the errors, that pattern obviously works. That works really, really well. So, yeah, coding agents that do that sort of error message loop thing, those are proven to work, and they're going to keep on getting better, and that's going to be great.

The research assistant agents are just beginning to get there. The things I'm critical of are the ones where you trust, you trust this thing to go out and act autonomously on your behalf and make decisions on your behalf, especially involving spending money. Like that, I don't see that working for a very long time. That feels to me like an AGI level problem. It's funny because I think Stripe actually released an agent toolkit, which is one of the things I featured that is trying to enable these agents each to have a wallet that they can go and spend and have basically, it's a virtual card. It's not that, not that difficult with modern infrastructure. If I can stick a $50 cap on it,

then at least it can't lose more than $50. You know, I don't know if either of you know Raphat Ali, he runs skiffed, which is a travel news vertical. And he constantly laughed at the fact that every agent thing is, we're going to get rid of booking a plane flight for you. And I would point out that historically, when the web started, the first thing everyone talked about is you can go online and book a trip, right? So it's funny for each generation of technological advance.

The thing they always want to kill is the travel agent. And now they want to kill the web page travel agent. I use Google Flight Search It's great Right if you gave me an agent to do that for me It would save me I mean maybe 15 seconds of typing in my things But I still want to see what my options are

And go yeah I'm not flying on that airline No matter how cheap they are Yeah For listeners Go ahead For listeners I think you know I think both of you are pretty positive

On Notebook LM And you know We actually interviewed the notebook LM Creators And they're actually two internal agents going on internally. The reason it takes so long is because they're running

an agent loop inside that is fairly autonomous which is kind of interesting. For a definition of agent loop if you picked that particular one. And you're talking about the podcast side of this, right? Yeah, the podcast side of things.

They have a there's going to be a new version coming out that will be featuring at our conference. That one's fascinating to me. Like Notebook LM, I think it's two products, right? On the one hand, it's actually a very good

rag product, right? You dump a bunch of things in, you can run searches that it does a good job of. And then they added the podcast thing as a bit of a, it's a total gimmick, right? But that gimmick got them attention because they had a great product that nobody paid any attention to at all. And then you add the unfeasibly good voice synthesis of the podcast. Like it's just spooly brilliant.

It's the lesson of like mid-journey and stuff like that. If you can create something that people can post on socials, like you don't have to lift a finger again to do any marketing for what. what you're doing. Let me, let me dig into Notebook LM just for a second as a podcaster. As a gimmick, it makes sense. And then obviously, you know, you dig into it. It sort of has problems around the edges. Like, it does the thing that all sort of LLMs kind of do where it's like, oh, we want to wrap up with a conclusion. I always call that like the eighth grade book report paper problem where it has to have an intro and, you know. But that's sort of a thing where,

because I think you spoke about this again in your piece at the year end, about how things are going multimodal and how things are that you didn't expect, like vision and especially audio. So that's another thing where, at least over the last year, there's been progress made that maybe you didn't think was coming as quick as it came. I don't know. I mean, a year ago, we had one really good visual model. We had GPT4 vision, was very impressive.

and Google Gemini had just dropped Gemini 1.0, which had vision, but nobody had really played with it yet. Like Google hadn't, people weren't taking Gemini seriously at that point. I feel like it was 1.5 Pro when it became apparent that actually they got over their hump and they were building really good models. And yeah, to be honest, the video models are mostly still using the same trick, the thing where you divide the video up into one image per second and you dump that all into the context. So maybe it shouldn't have been so surprising to us that long context models, plus vision meant that video was starting to be solved. Of course, not being, what you really want with videos,

you want to be able to do the audio and the images at the same time. And I think the models are beginning to do that now. Like originally, Gemini 1.5 Pro originally ignored the audio. It just did the, like, one frame per second video trick. As far as I can tell, the most recent ones are actually doing pure multimodal. But the things that opens up are just extraordinary. Like the chat GPT iPhone app feature that they shipped as one of their 12 days of OpenAI, I really can be having a conversation and just turn on my video camera and go,

hey, what kind of tree is this and so forth? And it works. And for all I know, that's just snapping a picture once a second and feeding it into the model. But the things that you can do with that as an end user are extraordinary. Like that to me, I don't think most people have cottoned onto the fact that you can now stream video directly into a model because it's only a few weeks old. But wow, that's a, that's a, that's a big boost in terms of what kinds of things you can do

with this stuff. Yeah. For people who are not that close, I think Gemini flashes free tier allows you to do something like capture a photo, one photo every second or a minute and leave it on 24-7 and you can prompt it to do whatever. And so you can effectively have your own camera app or monitoring app that you just prompt. And it detects for changes. It detects for, you know, alerts or anything like that.

Or it describes your day. You know, and the fact that this is free, I think it's also leads into the previous point of it being, the prices having come down a lot. Even if you're paying for this stuff, like a thing that I put in my blog entry is I ran a calculation and what would cost to process. process 68,000 photographs in my photo collection, and for each one, just generate a caption. And using Gemini 1.5 Flash 8b, it would cost me $1.68 to process 68,000 images, which is, I mean, that doesn't make sense. None of that makes sense.

Like, it's a 1,400th of a cent per image to generate captions now. So you can see why feeding in a day's worth of video just isn't even very expensive to process. Yeah, I'll tell you what is expensive. it's the other direction. So here we're talking about consuming video. And this year we also had a lot of progress. Like probably one of the most excited, excited, anticipated launches of the year was SORA.

We actually got SORA and less exciting. We did. And then V-O-2, Google's SORA, came out like three days later and upstaged it. Like Sora was exciting until V-O-2 landed, which was just better. In general, I feel the media or the social media has been very unfair to Sora, because what was released to the world generally available was Sora Light is the distilled version of Sora, right? I do not realize that. You're absolutely comparing the most cherry-picked version of VO2, the one that they published on the marketing page,

to the most embarrassing versions of Sora. So of course it's going to look bad. Well, I've got access to the VO2, I'm in the VO2 beta, and I've been poking around with it and getting it to generate pelicans on bicycles and stuff. I would absolutely believe that VOTU is actually better. Is Sora? So is full fat Sora coming soon? You know, when do we get to play with that one?

No one's mentioned anything. I think basically the strategy is let people play around with Sora Light and get info there. But keep developing Sora with the Hollywood studios. That's what they actually care about. Like the rest of us don't really know what to do with the video anyway. Right. I mean, that's my thing is I realize that for generative images and video, like images we've had for a few years.

and I don't feel like they've broken out into the talented artist community yet. Like lots of people are having fun with them and doing and producing stuff that's kind of cool to look at. But what I want, that movie, everything everywhere all at once, right? One ton of Oscars, utterly amazing film. The VFX team for that were five people, some of whom were watching YouTube videos to figure out what to do. My big question for Sora and Mid-Journey and stuff, what happens? when a creative team like that starts using these tools. I want the creative geniuses behind everything everywhere at once.

What are they going to be able to do with this stuff in like a few years' time? Because that's really exciting to me. That's where you take artists who are at the very peak of their game, give them these new capabilities and see what they can do with them. I should know a little bit here. So it should mention that that team actually used runway ML. So in that movie? Yeah.

I don't know how much. So, you know, it's possible to overstate. state this. But there are people integrating it generated a generative video within their workflow, even pre-sora. Right, because it's not the thing where it's like, okay, tomorrow we'll be able to do a full two-hour movie that you prompt with three sentences. It is like, for the very first part of, you know, video effects in film, it's like if you can get that three-second clip, if you can get that 20-second thing that they did in

the Matrix that blew everyone's minds and took a million dollars or whatever to do, Like, it's the little bits and pieces that they can fill in now that it's probably already there. Yeah, it's actually, like, I think actually having a layered view of what assets people need and, and letting AI fill in the low value assets, right? Like the background video, the background music, and, you know, sometimes the sound effects, that may be, maybe more palatable. maybe also changes the way that you evaluate the stuff that's coming out because people tend to, in social media, try to emphasize foreground stuff, main character stuff. So you really care about consistency and you really are bothered when, like,

for example, Sora botches an image generation of a gymnast doing flips, which is horrible. It's horrible. But for background crowds, like, who cares? And by the way, again, I was a film major way, way back in the day. Like, that's how it started like Braveheart where they filmed 10 people on a field

and then the computer could turn it into a thousand people on a field. Like, that's always been the way. It's around the margins and in the background that it first comes in. The Lord of the Rings, right? The Lord of the Rings movies were over 20 years ago. They had those giant battle sequences which were very early. Like, I mean, you could almost

call it a generative AI approach, right? They were using very sophisticated like algorithms to model out those different battles and all of that kind of stuff. Yeah, I know very little. I know basically nothing about film production, so I try not to commentate on it, but I am fascinated to see what happens when these tools start being used by the people at the top of their game.

I would say, like, there's a cultural war that is more being fought here than a technology war. Most of the Hollywood people are against any form of AI anyway. So they're busy fighting that battle instead of thinking about how to adopt it. And it's very fringe. I participated here in San Francisco, one generative AI video creative hackathon, where the AI positive artists actually met with technologists like myself, and then we collaborated together to build short films. And that was really nice. And I think I'll be hosting some of those in my events going forward. One thing that I think I want to give people a sense of is, like, this is a recap of last year, but then sometimes it's useful to walk away as well with, like, what can we expect in the future?

I don't know if you got anything. I would also call out that the Chinese models here have made a lot of progress. Hidua and Kling and God knows who else in the video arena. Also making a lot of progress. I think maybe actually China is surprisingly ahead with regards to open weights at least, but also just like specific forms of video generation. Wouldn't it be interesting if a film industry sprung up in a country that we don't normally think of having a really strong film industry that was using these tools.

Like that would be a fascinating sort of angle on this. Agreed? I am then. Oh, sorry. Go ahead. Just for people's, just to put it on people's radar as well, Hey Jen, there's a category of video avatar companies that don't specifically, don't specialize in general video. They only do talking heads, let's just say. And HAYgens didn't very well.

Swix, you know that that's what I've been using, right? Yeah, right. So if you see some of my recent YouTube videos and things like that, because the beauty part of the H-Gen thing is I don't want to use the robot voice. So I record the MP3 file for my clips every single day. And then I put that into H-Gen with the avatar that I've trained it on. And all it does is the lip sync. So it looks, it's not 100% uncanny valley beatable.

but it's good enough that if you weren't looking for it, it's just me sitting there doing one of my clips from the show. And yeah, so by the way, hey Jen, shout out to them. And so I would, you know, in terms of like the look ahead, going, like reviewing 2024, looking at trends for 2025, I would, they basically called this out. Meta tried to introduce AI influencers and failed horribly because they do just bad at it.

But at some point, there will be more and more, basically AI influencers, not in a way that Simon is, but in a way that they are not human. I, like, the few of those that have done well, I always feel like they're doing well because it's a gimmick, right? It's novel and fun to like that, the AI Seinfeld thing from last year, the Twitch stream. You know, like those, if you're the only one or one of just a few doing that, you'll get, you'll attract an audience because it's an interesting new thing. But I just, I don't know if that's going to be sustainable longer term or not. I'm going to tell you, because I've had discussions, I can't name the companies or whatever, but so think about the workflow for this. Like, now we all know that on TikTok and Instagram, like holding up a phone to your face and doing like in my car video or walking a walk and talk, you know, that's very common.

But also, if you want to do a professional sort of talking head video, you still have to sit in front of a camera. You still have to do the lighting. You still have to do the video editing versus if you can just record what I'm saying right now, the last 30 seconds, If you clip that out as an MP3 and you have a good enough avatar, then you can put that avatar in front of Times Square on a beach or whatever. So, like, again, for creators, the reason I think Simon were on the verge of something, it just, it's not going to, I think it's not, oh, we're going to have AI avatars take over. It'll be one of those things where it takes another piece of the workflow out and simplifies it. I'm all for that. I always love the stuff.

I like tools. tools that help human beings do more, do more ambitious things, I'm always in favor of. Like, that's what excites me about this entire field. Yeah, we're looking into basically creating one for my podcast. We have this guy, Charlie, he's Australian. He's not real, but he opens every show, and we're going to have him present all the shorts. Yeah, go ahead. The thing that I keep coming back to is this idea of credibility.

like in a world that is full of like AI generated everything and so forth, it becomes even more important that people find the sources of information that they trust and find people and find sources that are credible. And I feel like that's the one thing that LMs and AI can never have is credibility, right? Chat GPT can never stake its reputation on telling you something useful and interesting because that means nothing, right? It's a matrix multiplication. It depends on who prompted it and so forth.

So I'm always, and this is when I'm blogging as well, I'm always looking for, okay, who are the reliable people who will tell me useful, interesting information, who aren't just going to tell me whatever somebody's paying them to tell them, who aren't going to type a one sentence prompt into an LLM and spit out an essay and stick it online? And that, to me, like, earning that credibility is really important. That's why a lot of my ethics around the way that I publish are based on the idea that I want people to trust me. I want to do things that gain credibility in people's eyes so they will come to me for information as a trustworthy source. And it's the same for the sources that I'm consulting as well.

So I've been thinking a lot about that sort of credibility focus on this thing for a while now. Yeah. You can layer or structure credibility or decompose it. So one thing I would put in front of you, I'm not saying that you should agree with this or accept this at all, is that you can use AI to generate different variations. And then you pick you as the final sort of last mile person and you pick the last output. And you put your stamp of credibility behind that. everything's human reviewed instead of human origin.

That's the thing. If you publish something, you need to be able to be proud of publishing it. You need to say, I will put my name to this. I will attach my credibility to this thing. And if you're willing to do that, then that's great. For creators, this is huge because it's a fundamental asymmetry between starting with a blank slate versus choosing from five different variations. Right. And also, the key thing that you just said is like, if everything that I do, if all of the words were generated by an LLM, if the voice is generated by an LLM, if the video is also generated by the LLM,

then I haven't done anything, right? But if one or two of those, you take a shortcut, but it's still I'm willing to sign off on it. Like, I feel like that's where I feel like people are coming around to, like, this is maybe acceptable sort of. This is where I've been pushing the definition. I love the term slop,

where I've been pushing the definition of slop as AI generated content that is both unrequested and unreviewed, and the unreviewed thing is really important. Like, that's the thing that elevates something from slop to not slop is if a human being has reviewed it and said, you know what, this is actually worth other people's time. And again, I'm willing to attach my credibility to it and say, hey, this is worthwhile. It's the curational, curatorial and editorial part of it that, no matter what the tools are to do shortcuts, to do, as Swix is saying, choose between different edits or different cuts.

But in the end, if there's a curatorial mind or editorial mind behind it. Let me, I want to wedge this in before we start to close. One of the things, coming back to your year-end piece that has been something that I've been banging the drum about is when you're talking about LLMs getting harder to use. Oh, wow, yeah. You said most users are thrown in at the deep end. The default LLM chat UI is like taking brand new computer. users dropping them into a Linux terminal and expecting them to figure it all out. I mean, it's literally going back to the command line.

The command line was defeated by the GUI interface. And this is what I've been banging the drum about is like, this cannot be the user interface. What we have now cannot be the end result. Do you see any hints or seeds of a GUI moment for LLM interfaces? I mean, it has to happen. It absolutely has to happen. the usability of these things is turning into a bit of a crisis. And we are at least seeing some really interesting innovation in little directions.

Just like OpenAI's Chat GPT Canvas thing that they just launched, that is at least going a little bit more interesting than just chat chats and responses. You know, you can, exploring that space where you're collaborating with an LLM, you're both working on the same document. That makes a lot of sense to me. Like that feels really smart. One of the best things is still, who was it who did the, the UI where you could, they had a drawing UI where you draw an interface and click a button.

Yeah, TL Draw. TL Draw would then make it real thing? That was spectacular, absolutely spectacular, like alternative vision of how you'd interact with these models. Because, yeah, the, and that's, you know, so I feel like there is so much scope for innovation there, and it is beginning to happen. Like, I feel like most people do understand that we need to do better in terms of interfaces that both help explain what's going on and give people better tools for working with models. I was going to say, I want to dig a little deeper into this,

because think of the conceptual idea behind the GUI, which is instead of typing into a command line, open word.exe, you click an icon, right? So that's abstracting away sort of the, again, the programming stuff that, like, you know, a child can tap on an iPad and make a program open, right? But the problem, it seems to me right now with how we're interacting with LLMs is it's sort of like, you know, a dumb robot where it's like you poke it and it goes over here. But no, I want to go over here so you poke it this way and you can't get it exactly right. Like, what can we abstract away from the current what's going on that makes it more fine-tuned

and easier to get more precise. You see what I'm saying? Yes. And this is the other trend that I've been following from the last year, which I think is super interesting. It's the prompt-driven

UI development thing. Basically, this is the pattern where Claude artifacts was the first thing to do this really well. You type in a prompt, and it goes, oh, I should answer that by writing a custom HTML and JavaScript

application for you that does a certain thing. And when you think about that, and since then, it turns out, This is easy, right? Every decent LLM can produce HTML JavaScript that does something useful. So we've actually got this alternative way of interacting where they can respond to your prompt with an interactive custom interface that you can work with. People haven't quite wired those back up again.

Like ideally I'd want the LLM to be able to ask me a question where it builds me a custom little UI for that question and then it gets to see how I interacted with that. I don't know why that's like just such a small step from where we are right now. But that feels like such an obvious next step. Like an LLM, why should you just be communicating with text? When it can build interfaces on the fly that let you select a point on a map or move like sliders up and down. All of that. Nobs and dials. I keep saying knobs and dials.

We can do that. And the LLMs can build. And Claude artifacts will build you a knobs and dials interface. But at the moment, they haven't closed the loop. When you twiddle those knobs, Claude doesn't see what you were doing. they're going to close that loop. I'm a shock that they haven't done it yet. So yeah, I think there's so much scope for innovation and there's so much scope for doing interesting stuff with that model where the LLM, anything you can represent in HTML, JavaScript and SVG, which is almost everything, can now

be part of that ongoing conversation. Yeah, I would say the best executed version of this I've seen so far as Bolt, where you can literally type in, make a Spotify clone, make an Airbnb clone, and it actually does that for you zero shot with a nice design. Did you see there's a benchmark for that now? The L.M. Marina people now have a benchmark that is zero shot app, app generation, because all of the models can do it. Like, it's, I've started figuring out how I'm building my own version of this for my own project

because I think within six months, I think it'll just be an expected feature. Like if you have a web application, why don't you have a thing where, oh, look, the you can add a custom, like, so for my dataset, data exploration project, I want you to be able to do things like conjure up a dashboard, just via a prompt. You say, oh, I need a pie chart and the bar chart and put the next to each other, and then have a form where submitting the form inserts a row into my database table. And this is all suddenly feasible. It's, it's not even particularly difficult to do, which is utterly bizarre that these things are now easy. I think for a general audience, that is what I would highlight, that software creation is becoming easy and easy

here. Gemini is now available in Gmail and Google Sheets. I don't write my own Google Sheets formulas anymore. I just tell Gemini to do it. And so I think those are, I almost want to, basically somewhat disagree with your assertion that Ellen's got harder to use. Like, yes, we expose more capabilities, but they're in minor forms, like using Canvas, like web search in in chat, and like Gemini being in in Excel sheets or in Google sheets like yeah we're getting no no no no those are the things that make it harder because the problem is that for each of those features they're amazing if you understand the edges of the feature if you're like okay so in Google Gemini Excel formulas I can get it to do a certain amount of things but I can't get it to go and read a web you probably can't

get it to read a web page right but you know there are there are things that it can do and things that it can't do which completely undocumented if you ask it what it can and can't do they've terrible at answering questions about that. So, like, my favorite example is Claude Artifact. You can't build a Claude artifact that can hit an API somewhere else because the cause headers on that I-frame prevents accessing anything outside of CDNGS. So good luck learning cause headers as an end user in order to understand why. Like, I've seen people saying, oh, this is rubbish.

I tried building an artifact that would run a prompt and it couldn't because Claude didn't expose an API with course headers that all of this stuff is so weird and complicated. And yeah, like that, the more that with the more tools we add, the more expertise you need to really to understand the full scope of what you can do. And so it's, I wouldn't say it's, it's like that the question really comes down to what does it take to understand the full extent of what's possible. And honestly, that that's just getting more and more involved over time. Yeah. I have one more topic that I think you're kind of a champion of, and we've touched on it a little bit, which is local LLMs and running AI applications on your desktop. I feel like you are an early adopter

of many, many things. I had an interesting experience with that over the past year. Six months ago, I almost completely lost interest. And the reason is that six months ago, the best local models you could run, there was no point in using them at all because the best hosted models was so much better. Like, there was no point at which I'd choose to run a model on my laptop if I had API access to Claude 3.5 Sonnet. They just, they weren't even comparable. And that changed, basically in the past three months, as the local models had this step-changing capability, where now I can run some of these local models, and they're not as good as Claude 3.5 Sonnet, but they're not so far away that it's not worth me even using them. The other, the continuing problem is I've only

got 64 gigabytes of RAM, and if you run like Lama 370B, most of my RAM is gone. So now I have to shut down my Firefox tabs and my Chrome and my VS code windows in order to run it. But it's got me interested again. Like the efficiency improvements are such that now, if you were to like stick me in the desert island with my laptop, I'd be very productive using those local models. And that's, that's pretty exciting. And if those trends continue and also, like I think my next laptop, if, when I buy one is going to have twice the amount of RAM, at which point maybe I can run the, almost the top tier, like, open weights models

and still be able to use it as a computer as well. Nvidia just announced their $3,000, $128 gigabyte monstrosity. That's a pretty good price. You know, that's, um, that's, if, if that things is, customer OS and all. If I get a job, if, if, if, if I have enough of an income that I can justify blowing $3,000 on it, then yes. Let's do a GoFundMe to get Simon one of it.

Come on. You know you can get a job any time you want. This is just purely discretionary. I want a job that pays me to do exactly what I'm doing already and doesn't tell me what else to do. That's the challenge. I think Ethan Molek does pretty well. Whatever it is he's doing. But yeah, basically I was trying to bring in also, you know, not just local models,

but Apple Intelligence is on every Mac machine. and you're you see you skeptical. It's rubbish. Apple intelligence is so bad. It does one thing well. Oh yeah, what's that? It summarizes notifications, and sometimes it's humorous. Are you sure it does that well?

And also, by the way, the other, again, from a sort of a normie point of view, there's no indication from Apple of when to use it. Like, everybody upgrades their thing and it's like, okay, now you have Apple intelligence, and you never know when to use it ever again. Oh, yeah, you consult the Apple Docs, which, which is mkbhd. The one thing I'll say about Apple intelligence is one of the reasons it's so disappointing is that the models are just weak.

But now, like, Lama 3B is such a good model in a 2 gigabyte file. I think give Apple six months and hopefully they'll catch up to the state of the art under small models and then maybe it'll start being a lot more interesting. Yeah. Anyway, like, this was year one. And, you know, just like first year of iPhone, and maybe not that much of a hit. And then year three, they had the app store. So I would say give it some time.

And, you know, I think Chrome also shipping Gemini Nano, I think, this year in Chrome, which means that every app, every web app will have for free access to a local model that just ships in the browser, which is kind of interesting. And then I think I also wanted to just open the floor for any, like, you know, any of us, what are the apps that, you know, AI applications that we've adopted that have, that we really recommend. Because these are all, you know, apps that are running on a browser that, like, or apps that are running locally that we should be, that other people should be trying, right? Like, I feel like that's, that's one, it was one thing that is helpful at the start of the year. Okay.

So for running local models, my top picks, firstly, on the iPhone, there's this thing called MLC chat, which works, and it's easy to install, and it runs Lama 3B, and it's so much fun. Like, it's not necessarily a capable of novel. but I use it for real things. But my party trick right now is I get my phone to write a Netflix Christmas movie plot outline where like a jeweler falls in love with the King of Sweden or whatever. And it does a good job and it comes up with pun names for the movies.

And that's deeply entertaining. On my laptop, most recently I've been getting heavy into O-Lama because the O-Lama team are very, very good at finding the good models and packaging them up and making them work well. It gives you an API. My little LLM command line tool that has a lot. a plugin that talks to O-Lama, which works really well. So that's my O-Lama is, I think, the easiest on-ramp to running models locally.

If you want a nice user interface, LM Studio is, I think, the best user interface thing of that. It's not open source. It's good. It's worth playing with. The other one that I've been trained with recently, there's a thing called, what's it called OpenWeb UI or something? The UI is fantastic. Like, if you've got OLama running and you fire this thing up, it spots OLama and it gives you an interface onto your OLama models.

And that's really nicely done. That's my current favorite, like, open source UI for these things. But yeah, so there's lots of good options. You do need a lot of disk space. Like, the models are, the best, the models start at 2 gigabytes for, like, the 3B models that are actually worth playing with. The really impressive ones tend to be in the sort of 20 to 30 gigabyte range in my experience. Yeah. I think my struggle here is I'm not that much of an absolutist in terms of running things locally. Like, I'm happy to call an API. Same here. Okay, yeah, I just... I do it to play. It's my research interest, yeah. When people get so excited, answer your own question. Like, give us more apps that you want to... Yeah, sometimes it's like, it's just nice to recommend apps. So I use Super Whisper now. I try Whisper Flow. Didn't really work for me.

Super Whisper is one of them, which basically replaces typing. Like you should just talk most of the time, especially if you're doing anything long form. I hold down caps lock and I talk. And then when I'm done it, I lift it up. And it uses, it's not just about writing down your transcripts because I make ums and us all the time. I restate myself all the time. But it uses GPT4 to rewrite.

And that's what these guys are doing. They're all doing some form of state-of-the-art ASR automatic speech recognition. and then an LLLM to rewrite. And then I think I would also recommend for people to check out Rosebud for journaling. I think AI for Mental Health is quite unexplored. And it's not because we're trying to build AI therapists. I think the therapist really hate that. You'll never be on the level of therapists.

That gets back to the human thing that we were discussing. You know, on some level, there are certain things and disciplines that require the human touch. Sure. But the human touch costs me $300 an hour. Yes. Right? And this thing's $3 a month.

Like, you know. So there's a spectrum of people for whom that will work. And I think it's cheap now to try all these things. I'm going to turn a quick recommendation for an app. Mac Whisper is my favorite. That's your one. Desktop app. I love that thing.

Yeah, yeah, yeah. It runs Whisper. And you can do things like you can paste in the URL to a YouTube. video and it'll pull the audio and give you a transcript. So that's how I watch YouTube now. I slap it into Mac Whisper and then I hit copy and paste into Claude and then I use the Claude web app to do things. But Mac Whisper, it works with MP3 files.

Every time I'm on a podcast, I dump the MP3 into Mac Whisper, then I jump, dump the transcript into Claude and say, what should I put in the show notes? And it spits out a bullet point list where it says, oh, you mentioned like dataset that you should link to that, that kind of thing. Stuff like that. That's Mac Whisper. I use it several times. to be honest. Like it's great.

Yeah. I'm actually, I'm going to say one that is incredibly super basic and again coming back to just my workflow, but we are currently recording this on riverside.fm. Riverside is a great tool for recording video, audio, things like we're doing right now. But I always use this as an example to folks when they're like, well, how, what will AI do for me? When I first started using Riverside, like we're recording three different channels right now, right? You guys are recording locally. So there's three audio files, three video files. And then when I first started using Riverside, you had to pump three tracks into Adobe and then edit.

Okay, now we focus on Simon. Now we focus on Swix. Now we focus on Brine. Now we do all three. And then one day, a tool popped up that says, hit this button and it's smart edit. And then the AI determines, okay, Simon has been talking for 30 minutes. So go to the full shot of him. and Brian is now talking or there's over talk, so let's have all three talking heads.

With one button for anything I posted, it saved me three or four hours worth of work. That to me is like, again, if Normies are listening. Riverside has that feature now. Yeah. Yeah. I don't use it. Oh, that sounds fantastic. I still use a human editor.

The day it came out, I was running around the house telling my wife, telling anyone that would listen, you don't know. I just saved three hours because. they had a new feature. Like, that's, that's, Brian's basically crying, crying with joy right now. All right.

Let's, let's try to bring this to a landing a little bit. Simon, I have about, maybe two or three more. We can do these rapid fire. Cool. One of my shows,

one of the things of my show is, it's sort of like Silicon Valley writ large, so it's sort of like the horse race of who's up and who's down or whatever. To the degree that you're interested in pontificating on this, open AI, as a company in 2025. Do you see challenges coming?

Are you bearish, bullish? I almost am doing a CNBC sort of thing. But like, how do you feel about Open AI this year? I think they're in a bit of trouble. They seem to have lost a lot of talent. Like they're losing. And they don't have that. If it wasn't for 03,

they'd be in massive trouble because they'd have lost that, like, top of the pile thing. I think 03 clawed them back up again. But one of the big stories of 2024 is Open AIO started as the clear leader. and now Google Gemini is really good. Google Gemini had an amazing year. Anthropic Claude 3.570 is still my personal favorite model. And that feels notable.

Like Open AI went from, like, nobody would argue they were not the leader in all of this stuff a year ago. And today, they're still doing great, but they're not, like, as far ahead as they were. Next question, and maybe this couldn't be as rapid fire. But I loved finally from your piece the idea. that LLMs need better criticism, which I'd love you to expand on because as I sort of straddle this world of tech journalism and creator and investor and all that stuff, I thought that you had a really interesting thing to say about how, and we even alluded to this about like Hollywood being against it, like better criticism in the sense that as I took it, everybody is sort of,

they've got their hackles up, they're trying to defend their livelihoods and things like that. but it's either this is going to destroy my job and destroy the world or like I'm sorry I'm again leading the witness what did you mean by LLMs need a better criticism so this is a frustration I have that I like if I read a discussion thread somewhere about on this topic I can predict exactly what everyone's going to say people talk about the environmental impact they talk about the plagiarism of the training data the unlicensed training data they'll there's often this sort of oh and these things are completely useless thing that's the one that I will put it. push back against. The other things are true, right? The idea that LLMs are just completely useless,

the argument I will always make there is they are very useful if you understand how to use them, which is distinctly unintuitive. Like you have to learn how to deal with something that will just wildly hallucinate and make things up and all of those kinds of things. If you can learn how to, what they're good at and what they're bad at, I use them dozens of times a day and I get enormous value out of them. So I'll push back on people who say, no, they're just useless. But the other things, you know, the environmental impact of the way the training data works, I feel like the training data one's interesting because it's probably legal under fair use, but it's clearly unfair if somebody takes your work without your permission and trains a model

which then competes with you in the marketplace. Like legal or not, that's, I understand why people are upset about that. That's a reasonable thing to be upset by. So what I want, and I also feel like the impact that this stuff can have on society is especially as it starts undermining all sorts of jobs that we never thought were going to be undermined by technology. Like, who thought it would come for artists and lawyers first, right? That's bizarre. We need to have really high-quality conversations where we help people figure out what works, what doesn't work. We need people to be able to make good decisions about what to do with their careers to embrace this stuff and all of that sort of stuff. And if we just get distracted by saying, yeah, but it's useless plagiarism-driven, like, environmental, vent-ventically contrast, catastrophic. even though those things represent quite a lot of truth, I don't think that that's a useful message to lead with.

Like, I want to be having the much more interesting high-level conversations. Oh, okay, well, if there are negatives, what do we do to counter those negatives? If there are positives, how do we encourage those? How do we help people make good decisions about how to use this technology? I think where I see this the most is for people who are kind of very internal, know, like, sort of you and I are immersed in this every single day. So we're frankly tired of the same debates being recycled again and again. I think what might be more useful or more impactful is the level at which it starts to hit regulation. Last year, we had a couple of very notable

attempts at the White House level and in the California level to regulate AI. And those did not come to pass. But at some point, these criticisms bubble up to law, to matters of national security or national science in progress. And I feel like there needs to be more information or enlightenment there maybe, if only because it tends to be that they're very trailing. Like the, you know, my favorite example to pick on, which is very unfair of me, but whatever, you know, the California SB 1047 Act tried to cap compute at 10 to the power 25. Which is exactly deep sync. Exactly.

it also is exactly at the point at which we pivoted from training GPT5 to 01 where there is not no longer scaling pre-training compute. It's what I'm saying is like we're always trying to regulate the last war. And I don't think that works in a field that is basically eight years old. I think I've got there are two areas of regulation I'm super interested in that one of them is I do think that regulating the way these things are used can work. The big example is I don't want somebody's insurance claim denied by a black box LLM where nobody can explain what it did.

We have real laws for that. Exactly. Those laws. Take those laws, reinforce them, update them for modern capabilities. And then the other one, there's some really interesting stuff around privacy. Like we've got this huge problem right now where people will refuse to use any of these tools because they don't trust that the things they say to it won't be trained on and then exposed to other people.

And there are lots of terms and conditions that you can read through. and try and navigate around. I would love there to be just really straightforward laws that people understand where they know that it's not going to train on their input because there's a law that says under these circumstances, that can't happen. Like that sort of stuff, it's basically taking our existing privacy laws

and giving them a few more teeth and just reinforce them without introducing cookie banners a laeer the European Union, right? These things are always very, it's very risky to try and get the stuff right because you can have all sorts of, bad results if you don't design them correctly. But that's, there's space for that, I think.

Yeah, I, when I read that piece and then when you just said, you know, SWIC said, we were in the weeds on this every single day, so we're tired of hearing these arguments. It reminds me of folks that are always into politics. And then they're like, they're mad at the people that don't care about politics until it's an election year. and then they're like, well, you're a low information voter because all you know is that the factory in your town got shut down or there's inflation or whatever. And so you vote one way or the other, but you haven't been paying attention. But that's kind of the point is that you shouldn't expect normal people to pay attention, except for the fact that, oh, this might lose me my job. So you can't blame them for being, I don't know, for reactionary is the word or whatever, or emotional.

But right, if you're in the weeds, it's harder to keep everybody informed. And this is going to touch everybody. So I don't know. Okay. So this is the very last one. And then we can wrap and do plugs and everything. But Simon, this is for you. It was kind of alluded to a little bit.

And you might not have one. But if there's something this year that a generalist like me is not aware that is coming down the pike that you think is going to be big in the AI space, And maybe Sean, if you've got one too, what do you think it would be? I think for most people who haven't been paying attention, we know these things already. We know that the models are now almost free to run things against. The fact that you can now do video, like stream video to a model, the one that I've not played with nearly as much, but the thing where you can share your entire screen with a model and get feedback there,

that's going to be really useful. Like that's, again, the privacy side of things really matters, though. I do not want some model just training on everything that it's. sees on my screen. But no, that's, I feel like, like, the stuff that is now possible as of a few months ago is, is, that's enough. I don't need anything new. That's going to keep me busy all year. Swixie, you got one? Simon's always too content. And then, and then he sees the next thing and he's like, oh yeah, that's great too. Yep. Okay, I love trying to be contrarian by saying what does, what does everyone, what does everyone hate right now? And remember this time last year, we had, we just had

CES, Rabbit R1. We had the humane. Wearables, wearables, yep. Those are completely in the gutter. No one will touch them. They're toxic nuclear waste. Okay, this year is the year wearables. Yeah, yeah. I agree with you.

By the way, that cycle always works out where, like, you go to a CES and it's everything, hype, hype, hype, hype, and then three years later, it becomes the thing, unless it's 3D TVs, in which case that was a mistake anyway. But yeah. Transparent TVs are the big thing for the last couple of years.

What the hell? Yeah. Yeah. You know, so I think Simon may have got one of these, but there are a lot of people working on air wearables here in SF. They are surprisingly cheap, surprisingly capable, and with this a battery life, and they do useful things. We have to work out the privacy aspect, of course. But people like limitless, which used to be called rewind, I think, they're shipping, one of these wearables that based on your voice only records your voice. So you opt in. Interesting. Right.

Right. And so you can have perfect memory if you want. You can have perfect memory at work. Your employer can buy this for you that only, it only applies at work. And it's fine. It's just a meeting aid. Lots of people use granola or some kind of fireflies or like some of these meeting recorders only for for online meetings. But what about in person meetings? Or about conversations and locations that you've been? And some of that should be. a choice. Right now you have zero choice. You, and, and I think these wearables will enable some of that. And it's up to us as a society to determine what's acceptable and what's not.

I really like these gray areas where we still don't know yet. People, whenever I tell people about this, they're like, I don't know. Like, I'm sure, I guess it's like as though you have perfect memory, but some people have better memory than others. Like, where's the line? And there will be a lot more of these. I would add to that because SWIX, as you know, because you listen to my show, the idea that AI has taken the smart glasses and completely

changed everyone's mind about that as a product category and form factor. And I should say this, from things that I've been looking at investing in, wait till you see what they can add on to earbuds. Like the earbuds in your ear can do a lot more things than they're doing now. And then you combine that with smart glasses. and you combine that with an LLM that you can access, maybe with a phone as like the mothership. There's some interesting things.

CES next year is going to be crazy if you think wearables are AI wearables are a thing. Anyway, this year they were not a thing. There were very much no wearables as CES. This one's interesting as well because the thing that makes these interesting is multimodal, right? Audio input, video input, image input, which a year ago was hardly a thing

and now it's dirt cheap. So yeah, we're much better position now than we were 12 months ago to build the software behind this stuff. All right. Let's bring this to a landing. Swix go first. Tell everybody about obviously your podcast, which hopefully we're simulcasting, but also your conferences, events, everything. Sure, yeah. You can find my work on latent.

It's the AI engineer podcast, much more focused on serving engineers and developers. than the general audience, but you know, you feel free to dive in to the deep end with us. And we are also hosting a conference in New York in February, the AI Engineer Summit, where we gather people when, this one is entirely focused on agents. As much as, you know, people like to make fun of the idea that every year is the year of agents at work,

I think people at least want to gather to figure out what are the open problems to solve. And so these are the community of builders that get together, they show their latest work, I have Instacart coming to show how they use agents for their recommendation system and their sort of background jobs and internal jobs. And we have a whole bunch of like sort of financial tech company, FinTech or finance companies also showing off their work that I cannot name yet.

But it'll be lots of fun. We do high-quality events that sometimes people like Simon speak at. And that, right, as I said, or I think I said online or on air that I saw Simon speak at, one of your events last year. Wait, uh, Swix, just say again, it's in February, it's in New York City. I'm going to be there if that matters to anybody, if that's an attraction, but what's the dates on that and how to apply? Yeah, sir. I'm horrible at this. February 20th and 21st, is the leadership day for management, like VPs of AICTOs, and 21st is the engineer day, the individual contributors, hands and keyboard people. And that's when I'll have the big labs.

So deep mind, anthropic, meta, open AI, all coming to share their agents work. and then we'll have some new launches as well that you haven't heard of. And to sign up to attend, what website can I go to? Yeah, it's apply.a.ai.engineer. All right, Simon, I'm going to hold you or handhold you even more. Your weblog is Simon Willisson.net. But what else would you like us to know or go find out about what you're doing? Yeah, I was going to say my blog.

My day job, I call it a job, is I work on open source tools for data. journalism. That's my project. Datasets, spelled like the word cassette, but data. Dataset.com. And that's beginning to grow some interesting AI tools. Like, originally it was all about data publishing and exploration and analysis. And now I'm like, okay, well, what plug-ins for that can I build that you let you use LMs to craft queries and build dashboards and all sorts of bits and pieces like that? So I'm expecting to have some really interesting product features along those lines in the next few months. And I'll end by saying, if anyone everyone's listening to this on Swix's show.

I do the TechMeme Right Home. Every single weekday, 15-minute long, tech news podcasts, look up Ride Home on your podcast app at Choice. TechMeme Right Home. Gentlemen, thank you for your time. Thank you. This was fantastic. What a great way to start the year for this show.

Cool. Thanks a lot for having me. This has been really fun. Yeah, thanks for having us. I'm honored to be on.