Inside Google's AI Slump: Can Gemini 4 Argon Fix It?

Ejaaz When it comes to AI, Google is a company that you would expect to see at the top. And yet over the last six months, they've suffered a bit of a fall from grace. Their flagship model, Gemini, which peaked at the end of last year, has suffered delay after delay after delay, with companies like Anthropic, OpenAIR, and even Meta running complete circles around them.

Ejaaz I think 15 brand new AI models has dropped since Google's last one has. But the bad news doesn't stop there. Google has also recently lost key members of staff from their AI team, the most prominent one being Demis Hacibis, the AI CEO of DeepMind, who started the whole thing at Google. He's had to step away to focus on other things, but they also lost their chief AI scientist, Jeff Dean, who, for all intents and purposes, is the godfather of Google.

Ejaaz He's responsible and the man behind 90% of Google's hit products, including the TPUs that power Gemini and power other models like Claude and ChatGPT. As well as having his hands in several other of foundational research papers that has given birth to modern day AI as we know it. That guy also left citing reasons that he can't do the work that he wants to work on at Google anymore. And so the public reaction was understandably catastrophic.

Ejaaz The stock dropped 5% over a matter of hours, losing the company tens of billions of dollars. But recently, I think things have started to shift. Now, of course, Google in itself, on its own, without all the AI stuff, is a massively successful company. They're the third most valuable company in the world. I'm not negating that.

Ejaaz But if AI is set to be the future, it's important that a company is able to pivot. When I look at the likes of Meta, when I look at the likes of Amazon, although they're operating in different spheres, they are able to focus and spend so much money and time and resources in trying to figure out the AI thing.

Ejaaz And Google started very much in the same way. They've been arguably the longest company in this game spanning back decades and decades. And so after this recent shift in staff, time and resources, it's important that they're able to get their foot back into the race. And we might be starting to see it.

Ejaaz Just last week, they released Gemini 4 Argon, or at least they announced it. And if benchmarks are anything to go by, this model is pretty powerful and can compete with the best of the best from Anthropic and OpenAI. But they also recently, in fact, today, as I'm recording this, released Nano Banana 2.1, their brand new image generation model.

Ejaaz So we're starting to see more releases from Google. And on this episode, I want to dig into what's happened, where Google is at right now, and what the future might look like. All of this and more on today's episode of In The Loop, brought to you by Qualcomm. Now, very quickly, before we get into it, you might be wondering, Ejaz, this is a different setup and you're missing a co-host.

Ejaaz There's a ton of updates that have happened for the show. Go and check it out on our video that we posted two days ago. It's on our channel and we'll explain everything there. But long story short is Limitless is evolving into a new brand refresh called In The Loop, but it's very much the same DNA of the show. And we're going to be coming to you every single day of the week with the latest and greatest in AI and frontier tech.

Ejaaz All right. So now setting the stage for Google as a company. Now, for the last seven months, Google has basically had their best model underperform every single flagship product that Anthropic and OpenArea have put up. And I mean that for every single model, including Opus, including Sonnet, which are typically the cheaper, less impressive tiers from Anthropic, including Sol, including Luna and Terra, which again, are the lowest equivalents coming from open air, their cheapest and technically worst models would still outperform Google's best.

Ejaaz And that's weird, because at the end of last year, Gemini was on top and they had a meteoric comeback where they kind of had the lead back in the day before GPT came out. And then GPT came out and took the world by storm. And Google was criticized for not being able to. Stage a comeback and actually catch up with all the research that they've done.

Ejaaz And then they released Gemini and the world's opinion changed on it. It became deeply integrated into every Google product that you could use, including Google search. And it just became synonymous with using Cloud or ChatGPT. But something changed since then. And I don't know whether it's a research problem, or whether it's a staffing problem, or whether it's just generally a product company focus thing, but they weren't able to ship a competent model.

Ejaaz Now, speculation behind that is at the same time that they started to announce delays, Claude Code hit an absolute exponential run. And this, of course, is the flagship product from Anthropic. And at that time, AI coding was kind of this mysterious, weird thing. Companies didn't really know whether they should focus on automating coding or not, whether they should focus on automating research or other such things.

Ejaaz And Claude was the very first instance where it became clear coding and AI are a natural fit. And that is where the leading edge is going to be figured out. OpenAI figured this out about a month after Claude Code launched and we shifted everything to build their own competitor called Codex. And since then, they've been able to catch up. And you could argue that's why they've been able to build much better models that stay on par with Claude.

Ejaaz And now it's Anthropic versus OpenAI, because they're using these models to build the next better versions of themselves, something called recursive self improving. Google, People unfortunately watched all of this happen and thought. I think we can wait a bit. We can try and figure out all these other types of products that we've already committed to on our product roadmap and we can come back to coding.

Ejaaz And that ended up costing them pretty greatly, which is a shame because Google has some of the biggest distribution on earth. You're seeing on the screen here, their market share value is around $4.1 trillion, making them I think the third most valuable company in the world. And they have around 3 billion users across all their different products.

Ejaaz It's very important to state that there is no other company like Google that has such a large product surface area across both general consumers, so the public like you and I, and enterprise customers that use their products pretty aggressively to document stuff on Google Docs to create slides or to manage their databases via Google Sheets.

Ejaaz So pretty much wherever you are on the internet, whether you're searching for something, whether you're trying to produce something, whether you're trying to analyze something, there's always a very competitive google alternative which is why then when ai exploded onto the scene you would expect them to be able to pull off a similar stunt and they actually were able to with gemini.

Ejaaz But they haven't been able to keep up but don't take my word for it i mean look at the screen when we look at the artificial analysis index which is a really good general benchmark to try and test the intelligence of the latest model from these frontier companies you see clawed opus 5.5 which was released about a week and a half ago score 58 you've got gpt6 astra scoring a very close 53.

Ejaaz And then you have Gemini 3.1 Pro, which was released back in February, which in AI years, funny I say years, which in AI timeline is years and years and years worth of time. So you wonder why hasn't Google been able to ship an effective product? Well, the answer is, I think there is a work culture problem because starting August 5th, but actually even before that, they started to bleed a lot of very important talent. And we covered this on previous episodes.

Ejaaz We mentioned that John Jumper, who shared a Nobel Prize with Demisus for building AlphaFold, which is basically their protein sequencing DNA AI model, had left to join OpenAI. And then Jeff Dean, who I mentioned earlier on in the intro, who is kind of like the chief scientist and grandfather at Google for a lot of the other products that aren't just AI related, just an absolute invaluable person at the company, left to start his own AI lab.

Ejaaz I think it's called Discovery Loop, which is focused on automating research, first starting in AI, and then other parts of research in physics and science. That is exactly what the Frontier AI labs that are leading this entire field are doing and focusing on. So my question then becomes, why on earth did a loyalist like Jeff Dean feel the need to leave when he could have easily built this at Google?

Ejaaz I think there's a work culture problem going on. So it led to this massive exodus. Demis' service had to step down and focus on something else. And so what we're left with is a completely bare team that needs to figure out a completely new strategy. But everything I'm saying, I admit, sounds very Doomer-esque for a company that is the third most valuable in the world and is making so much money.

Ejaaz Because when you look at the stats, right, they have over 1 billion in AI mode for search past a billion in AI overviews, which have reached around 2.5 billion. And then when you look at the amount of tokens that they process for their AI products or across all of their products, it's 22 billion tokens a minute.

Ejaaz They've also reported 82% growth in Google Cloud, which is their AI CapEx infrastructure business where they're spinning up TPUs, which is their own version of NVIDIA's GPUs, and using it to facilitate selling training and inference costs to other AI labs or firms that want to train and raise their own models.

Ejaaz And again, putting all of that aside, they have a humongous enterprise business, 8 million Gemini enterprise paid seats on their AI products alone. That's not even including some of their Google Cloud enterprise products. So for all intents and purposes, is Google is an absolute cash-generating machine, and I'm not negating that. They have one of the largest cash balances on Earth for any company.

Ejaaz But they've also been spending very aggressively to try and acquire and build as much compute data centers to pay for the expensive capex costs that they're taking. But there's a bit of a mismatch. Google right now, as they're positioned, is focused to be a really good hyperscaler or infrastructure provider, akin to maybe a neocloud that builds AI data centers or builds the chips, supplies them then to AI labs that can use them for inference costs or training their own AI models.

Ejaaz And that is all fair and game. But Google at its core DNA is not just a research lab. They're this perfect hybrid of a research lab and a product company. That's why they've been able to become one of the biggest companies on Earth. And that's why people love Google as a company and as a business. They haven't been able to translate this for AI yet.

Ejaaz And you can argue that if they maintain this kind of product or working culture, then they're going to end up just remaining a hyperscaler or a neocloud, whilst companies that actually build the net new products like AI models, like AI agents that we've spoken about on previous episodes, will end up benefiting from the economic opportunity that just explodes over the next couple of decades.

Ejaaz So with that said, now has never been a more important time for Google to shift gears and try and figure out an alternative strategy to get back in the AI race. I personally want to see Google win because they have such a huge advantage. They're not just a singular monolithic company. They have fingers in every different pie that you could think of when it comes to the internet.

Ejaaz They're the most vertically integrated company that doesn't exist for any other AI lab. When you think of OpenAI and Anthropic, yes, fantastic models, but they lack the distribution. They don't have the billions of users. When you look at Meta, they have all the billions of users, but they don't have a bleeding edge model.

Ejaaz Google, with all their wealth, resources, and expertise, should be able to pull off a miracle and have the best vertically integrated AI product. They have all the data to kind of mold that into the best consumer experience or enterprise experience depending what kind of a product they want to build.

Ejaaz But the result of what we've actually seen is very fractured teams being divvied up, very specific resources that don't really make sense. This team gets compute, but the other team doesn't get compute. Or another team gets way more compute, but it's spent on a product that maybe may not have as much impact on millions of users versus a couple hundred thousand.

Ejaaz It's just this weird thing. And then they're selling compute to competitors when they probably should be using it to train their own model just to make cash in the near term so that it might look good on a quarterly earnings report versus trying to play the longer game. And so the question then becomes, okay, what moves will Google make under this new realm of losing half the AI talent and trying to figure out a new structure and approach?

Ejaaz Well, we might see it with a brand new model. They announced Gemini 4 Argon, which as the Google post says, delivers frontier performance in complex workflows across real world software engineering, knowledge work, and cybersecurity defense. Now, of course, The first thing that I looked at that Sundar Pichai, the CEO, highlighted was their benchmark evaluations.

Ejaaz Now, if you're watching the show, if you've watched it for a while, you know we're not big benchmark guys, but it gives us an idea of what this model might be. And so the first thing that draws my eye is, okay, it's beating GPT-6 Astra and Claude Fable 5.1, which are each respective company's flagship model model.

Ejaaz Pretty well on a number of different things. But then you have to look at what that thing is. Now on the left here top, it's knowledge work talking about VALS index automation bench knowledge work is basically if you think about 70 to 80% of the tasks that are required for a software based job, and I'm not talking about software engineering, but I'm talking about anyone that has a nine to five that uses a computer that uses slack and a bunch of these other different CRM tools or whatever that might be, you might need to access a spreadsheet or write up a strategy document or type messages to different workers in Slack or customer support, this model excels across any other frontier competitor model, which is fantastic,

Ejaaz which honestly applies to a much larger percentage of the world than some of these other benchmarks and what they test for. But moving on to agentic coding, which is kind of like the creme de la creme of benchmark tests, it excels in DeepSuite version 1.1, which is effectively the AI agent version of coding.

Ejaaz There's two types or different ways of testing an AI model and how good it is at coding. There's the monolithic version where you just ask an AI model itself to try and figure out a complex coding task. And then there's a version where you can allow that AI to spin up multiple versions of itself and do what is known as agentic reasoning, which is work between multiple versions of itself to try and figure out a solution to the problem.

Ejaaz DeepSuite is exactly that. And when you look at. Fable 5.1 and gemini 6.1 astra very impressive and then across other things like science math long context computer use it does a very good job especially when it comes to multimodal reasoning so overall it sounds like this has been a major upgrade for google in general and if i pull up this chart over here we look at the artificial analysis index which i mentioned earlier google had been trading behind by like 25 points they have now jumped up to 53 which matches gpt6 astra So for all intents and purposes, this is a good model, but I can't give you a genuine review because.

Ejaaz I haven't been able to get my hands on it. And that's because they haven't actually released it to the public for fear of causing a cybersecurity incident. Now, this has kind of become a recurring case across every AI lab that releases a new model, they train this model, and then they realize, oh, this thing could be incredibly misaligned, and could break out of a sandbox and cause a lot of disruption.

Ejaaz So let me sandbox and test it with cybersecurity defenders first, let them get their hands on it first. I think the very first model that did this was Claude Mythos Preview back at the start of the year, which is ironic because we now have a bunch of open source models which are as capable as that. And we haven't quite seen any of them cause AI disruptions, at least any that we have been aware of. But the point still stands. We want to build and release AI effectively and safely, which I'm fully on board with.

Ejaaz So Google is undergoing a process that I think was self-volunteered by the government, which I would guess lasts about one to maybe two weeks. So we should expect to see the public release sometime around a week from now. But some insiders as reported by Bloomberg has already got their hands on it. And listen, there's mixed reviews.

Ejaaz One person goes, it does less well when employees actually put it to work. The model struggles to handle certain coding tasks. And this has been my whole thing around Google and Gemini models. And this is just a personal take, you might have a different one. But once you get your hand on the model and actually end up using it, it is a stark and very different experience to what the benchmarks actually indicate to you. So the benchmarks look very impressive in theory.

Ejaaz But when you actually end up using it, you realize I'll just continue using GPT or continue using claw. Now I'll reserve my judgment until I actually get my hands on Gemini 4. For all intents and purposes, I'm excited to see Google come back in a very meaningful way. And especially in coding where they've honestly lagged behind for every other model iteration that they've had over the past year. So I'm remaining cautiously optimistic that they've been able to figure something out. All of this and more on today's episode of In The Loop brought to you by Qualcomm.

Ejaaz But that's not all that they're releasing. They also released a brand new version of their image generation model, which is Nano Banana 2.1. Now, this is brand new. It released literally a few hours ago. So I'm not gonna sit here and pretend like I have access to a bunch of different demos because I haven't tried the thing out itself.

Ejaaz But from the visuals and examples that we've seen so far, let me tell you, social media is cooked. These things are looking incredibly realistic. I'm pretty certain that there's no ladybug that has ever looked like this or creature or insect that has ever looked like this. Weird pink lucid neon bug would be kind of cool if it existed, but you know the humans are looking more human the reflections are looking way more accurate everyone has all their fingers in all the images there's no extra limbs or appendages added which is fantastic these models are getting really good and google has been in the game for multi-modality and by that i mean good image generation models great video models and fantastic world models i believe genie 3 we haven't seen

Ejaaz an iteration since then, but they were leading world model front as well. The point is, Google is perfectly positioned to be putting out some of these products, and they've been slow or delayed to release these products over the last six months. But we're starting to see a comeback since then, and finally, a shift in product and work culture.

Ejaaz Now, when I step back, and I think, okay, listen, from what we've seen, from the evidence over the last year, six months, and these products that have just released over the last week, how do we feel about Google? I stand by what I said earlier, which is they are a fantastic company that has so many different ways of making money. I think Google...

Ejaaz For most companies, has an incredible privilege of having years and years of runway ahead of them. But it's only years. And a lot can change in those years. And we've seen the likes of Anthropic and OpenAI increase their ARR pretty stupendously over the last 12 months. We've seen Anthropic go from their expected forecast for this year was 10 billion, and they've recently reportedly hit 100.

Ejaaz And OpenAI has also recently hit 70 billion ARR as well, the fastest growing companies that we've ever seen. If they continue that trajectory, and that's a big if, they should be able to catch up to Google's ARR in roughly about two and a half years. So Google has roughly that amount of time to figure their AI strategy out.

Ejaaz And it's not just Google that's alone on this. You can point to companies like Apple that have been very slow because they can afford to take the time. What I applaud Google for is that they have taken the time to test out different products in public. And maybe they've fallen flat on their face sometimes, but they can build and iterate from that. So it's going to come down to the team, it's going to come down with how they spend their resources, and how they shift in reaction to losing half of their most important AI talent ever. Jeff Dean.

Ejaaz Sucks to see you go, dude. And the final point I'll make is don't forget, Google is, again, the only major company that can vertically integrate the entire AI stack. They have the TPUs, so the GPU equivalent of training and inferencing AI. Then they have the AI models themselves. Then they have the distribution across billions and billions of users across, so many different interfaces, Google Maps, Google Search, Google Ads.

Ejaaz They can just apply AI and change it overnight. So I still think the ball is very much in Google's court and I think they need to start acting much quicker than what we're seeing and what gives me hope is for the longest time for those of you especially who have watched this show I, have personally been bearish on a company like Meta and I never thought that they'd be able to pull it around and, they surprised me and they've come out with one of my favorite products this year in the form of Metamuse, a personal AI agent I think Google can pull off a similar stunt but it remains to be seen whether they actually want to and whether they will actually commit resources to be able to do that.

Ejaaz That's basically it. That's my take on the whole Google thing. I wanted to give an update on Google pretty desperately because I miss it and I'm still very bullish on the company, but I've been struggling to try and figure out how they're able to turn things around. I've been waiting for a flagship model and maybe that's come in the form of Gemini 4 Argon.

Ejaaz I can't wait to get my hands on it. If you're listening to this and you are from Google, please, I would love to play around with it and give my honest review. But that is that. It has been a crazy couple of weeks, guys. This is a brand new version of the show. And the idea here is that we keep the core DNA of everything that we've done here at Limitless. And we bring the best, most insightful takes and thoughts on all the top AI trends and news that is happening week in, week out.

Ejaaz So if you enjoyed this episode and you aren't subscribed or you haven't left a comment, please do both of those things. In fact, do me a favor and even turn on notifications as well. It helps us out massively in terms of getting to a wider reach and audience and sharing as many opinions as we can. If you have a take that disagrees with me, I would love to hear from you genuinely.

Ejaaz Like I am always trying to pick holes in my arguments and in my takes and in my opinions. But at the end of the day, I'm the same as you. I'm trying to figure all of this out in real time. So I actually learn a lot from the comments that you guys give. And if there's anything that you can teach me or if there's any alternative perspectives that you might have that I might have missed that compliment some of the stuff that I've talked about, let me know.

Ejaaz If there is something or a different angle about Google that I've missed, I want to hear about it. You can DM me or you can leave a comment. I don't really care. I'll read anything. Finally, if you think a friend of yours or someone in your network might benefit from an episode like this, please copy the link and share it with them.

Ejaaz I want to get this to as many people as we can. We are also on Spotify, Apple Music, and pretty much wherever you can listen and hear a show or podcast. And that is it. Thank you so much for listening to In The Loop brought to you by Qualcomm. I'll see you on the next one. See you guys.

Inside Google's AI Slump: Can Gemini 4 Argon Fix It?

Kapitelklicken springt direkt hin

Shownotes

Transkriptvom Podcast bereitgestellt · Zeitmarken springen hin