Google & Claude: The AI hype trains are taking different tracks!
13 minAudio
What a week in AI? Well, okay, I guess every week is "packed with announcements". But with Google IO, tons of projects and the Claude 4 model releases, there WAS a lot going on. Time to unpack!
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
I'm recording this at the end of a week that was packed with AI-related launches. And you're right, every week is packed with AI-related launches. It can be rather annoying, but this week we had Google I.O., and therefore Google gave us tons of AI stuff, and we'll dive into that, and we had the Anthropic event where they launched their Claude 4 models, and that was also pretty interesting. Also because we can clearly see that
these companies are following totally different strategies, and we can see that the models developed by these companies focus on totally different things. And in general, we might therefore see that we maybe don't get AGI, general intelligence, through those LLMs anytime soon. Instead, we might get a lot of specialized products and models, and that's a rather interesting future, and in my opinion, the more
likely future than having that, uh, all-knowing super AI. But we'll see. And I actually wanna start with Anthropic and Claude 4 because when I'm recording this, it was just yesterday that they launched this all. And they launched two new models, two new versions of the Claude 4 model, so to say, and if you watched their livestream or if you read their announcement blog post, which of course, like many other sources, is linked below this episode, then you'll quickly
notice that they're only talking about coding. Their model is only about coding, or their models are only about coding, and that seems to be the niche. It's not really a niche, I guess, it's one of the things these, uh, large language models can do best, but it's their thing they're specializing on clearly. And it was kind of obvious, I guess, because with Claude 3.5 and 3.7, though that model
had some problems, but with those older models that were released a couple of weeks or months ago, we could already see that they were primarily strong in generating code. And in general, these large language models are decent at generating code, at least up to a certain point, and I'll also get back to that. So it's no surprise that Claude 4 is all about code and that they only mention (laughs) that in their announcement essentially, and their Claude 4 seems to do really well. Now, I
personally haven't used it a lot. I played around with it a bit. It's available through Cursor and VS Code, for example, but not so much in many other tools, and we'll see how well it does because I do remember that when Claude 3.7 was announced, it also was super amazing and all these AI influencers on X hyped it up like they do with every model, of course, and then we all found out, well, it's not that great. It kind of does random stuff, it does stuff you don't want it to do, it changes random parts of your code, and we'll have to
see if Claude 4 is any better there. But what's interesting is that Anthropic is clearly becoming that code large language model company. They want to win the code generation game, which is a difficult game to win because it's not like they're the only model creator that wants to win there, but it seems to be their thing. And of course, the general question is what's the end game for large language models and code generation? You could of course
argue that in a year or so, no code or almost no code will be written by human developers, and there have been some CEOs that followed that reasoning or that shared, uh, that idea, that, uh, way of thinking. I don't agree at all, and I think it's total nonsense to assume that this will be the future anytime soon. What's clear, of course, is that these models can be great assistants, and I've said that for, for months now.
I've been using these tools, like probably all of you, for months now as assistants, no matter if that's GitHub Copilot and the GitHub Copilot Agent mode in VS Code or if it's Cursor or Windsurf or any other extension or IDE. These models can be really good at that. By the way, not just when you're using them integrated into an IDE or code editor, but also as a partner for having discussions and for evaluating different approaches or architectures when you're planning a project, when you're setting up a more
complex project. That's what I'm doing a lot, uh, I'm using Google Gemini 2.5 Pro all the time for discussing different strategies, different ways of implementing certain features in a Go application I'm currently building which I'll hopefully launch soon. So I'm doing that all the time, and that's where they're really, really valuable, and of course, that will stay that way in the future and they will probably become more valuable there. But when it comes to generating all the code and
building more complex applications and not getting lost along the way, when it comes to these things, sorry, that's, that's just not going to happen anytime soon, that's totally nonsense in, in my opinion. But of course, for building individual features with help of AI, they're amazing and that's going to stay that way and you should use them for that, in my opinion, uh, you should definitely at least give them a try. And I've mentioned it before, I got a course on how to efficiently use, uh, AI tools as a developer if you wanna learn
more about that, I'll link it below the video. I also got some other AI courses which you'll also find there, but no matter if you take the course or not, give those AI models a shot.... for using them, uh, as, as assistants when writing code. And that seems to be the Claude thing, though of course they clearly, uh, have a different goal. They wanna replace developers in the end, I guess, because they definitely wanna, uh, keep on pushing and build models that are capable of building entire projects on their own or with
almost no human input, I guess. And we'll see how that goes, uh, and we'll have to see how good Claude 4 is. But yeah, here it is and Anthropic is clearly the coding large language model company for now and we'll see if that changes again in the future, but that seems to be their thing. Google, on the other hand, is totally different. They also had their keynote and lots of other sessions at Google I/O this weekend. It was all about AI, as you might imagine, and they announced lots of
initiatives, research projects, a new super expensive paid plan that gives you early access or exclusive access to certain models and features, uh, if you're in the US. Uh, us peasants from other countries are not included for now, but if you're in the US and you're willing to pay 250 bucks per month, after the first three months where it's a bit cheaper, uh, then you get all that extra access. And clearly Google is
trying to achieve a more general form of intelligence. They're going all in on agents. They announced a couple of research projects, specifically Astra and Marina, that are all about giving you, uh, agentic features no matter if it's integrated into Gemini or if it's on Google devices, like their Google smartphones, the Pixel devices, uh, which should be capable of helping you on all kinds of tasks without a lot of input from you. So they're definitely going all in there.
They announced View 3, uh, a video generation model which looked pretty amazing. Imagen-4, their image generation, uh, model which, uh, should be pretty good, though (laughs) I just tested it before recording this and it failed to create an illustration, but what can you do? Uh, maybe it was also the old model. I'm not even sure which model I'm using there, to be fair. But yeah, they announced a lot and I will say the Google Gemini models, the 2.5 models, are pretty amazing. I'm using them all the time, as I mentioned before, for discussing
how to implement certain features, how to tackle certain problems. And of course also outside of coding, because unlike Anthropic, Google clearly does not have a goal of only winning the coding game, though like all AI companies, it is a lot about coding and code assistance and coding agents, because that's like the quick win, right? Large language models ultimately are about generating text. Code is text, so that's there for a natural fit and where a lot of work flows into.
But it's not just about that. As you can tell by the benchmark numbers Google publishes, it's not all about code. Um, and about... In general, their messaging. It's about building those agents and about helping you with all kinds of tasks. Of course, Google being Google, they're also trying to integrate AI even deeper into their search. Ultimately, search may just turn into a chat in the end, like an AI chat where you can still search for
things, where you can still ask your questions, but then you'll get a, a result, an answer, which of course will include links to sources, but you will not primarily get those links as Google worked in the past and still does to some extent today. Instead you will get these summaries. And that of course has interesting implications for people creating content, writing news articles, blog posts, you name it. Because if you're just feeding the Google AI or other AIs for
their summaries and people never visit your website, what's the point of writing blog posts and news articles in the first place, right? So that will be an interesting future. Um, ultimately, we may have AI summaries of content that was also generated by AI and I'm not sure if I'm liking that future. But anyways, that's where we're going, and that is what Google also announced. They're adding more AI integration into Google Search. They are adding this AI mode there where you essentially have this chat mode. Um, so an, an
open, uh, AI chat without the need for a sign-up and anything like that. And it of course makes sense because they can't just let, uh, OpenAI and other providers eat their lunch with their models, but of course it's an interesting challenge for Google, and ultimately as mentioned, also for content creators. Now, of course we could dive into all the details of the different capabilities and projects and models that were announced. You probably already heard a lot about them, and it'll
be interesting to see which of these projects become reality, anyways, because announcing a lot of stuff is one thing. Having good products in the end is a totally different thing, and that's been the problem with AI for the last two years. There's been a lot of hype. There have been a lot of announcements. And don't get me wrong, there have been a lot of great products and services and capabilities and we're really living in an amazing or a scary world, depending on how you look at it,
when it comes to AI, but there also have been a lot of broken promises. I've heard lots of CEOs tell me that, uh, developers will be replaced in six months, 12 months, and that was two years ago. So lots of things being said. You should not take it all that seriously, but of course it'll be interesting to see, uh, where we'll be in a couple of months from now. And we definitely got interesting models, great coding assistants as mentioned, great
models for doing research and all that fun stuff. And yeah, that's essentially what happened, I guess not just this week but what also kept on evolving this weekend. Therefore, please definitely let me also know what your thoughts are about this entire development, about this week specifically, AI in general, and what you think where we'll be in, let's say, six months from now. I'll of course keep you updated with my thoughts and so on, and um, wish you a great time. Bye.
GPT 6 Astra was released. It's a good model, I have to say. Especially for 3D work it seems to be a real game changer. But due to all the hype, you may be disappointed...