Limitless: An AI Podcast - Google Took a Bite Out of Apple: The Gemini Masterplan

Episode Date: January 14, 2026

Google has had a big week, with $1 billion annual deal with Apple to integrate its Gemini AI across 2 billion devices, signaling a major shift in the infamous Siri. We also discuss Google’s market cap surpassing $4 trillion, and the launch of Gemini in Gmail, which automates email management. Additionally, we analyze Google’s Universal Commerce Protocol, aiming to redefine online shopping with partners like Walmart and Shopify.------🌌 LIMITLESS HQ: LISTEN & FOLLOW HERE ⬇️https://limitless.bankless.com/https://x.com/LimitlessFT------TIMESTAMPS0:00 The Week Google Changed Everything0:56 Google's Dominance in AI2:29 The Apple and Google Partnership3:01 Apple's Shift to Google AI7:05 The Future of E-Commerce17:38 Universal Commerce Protocol23:10 Google’s Strategy for 202624:42 Conclusion: A Big Win for Google------RESOURCESJosh: https://x.com/JoshKaleEjaaz: https://x.com/cryptopunk7213------Not financial or tax advice. See our investment disclosures here:https://www.bankless.com/disclosures⁠

Get this. In the span of just one week, Google got Apple to admit that their tech is superior and pay them $1 billion for the privilege of putting Gemini in every single Apple device. That's two billion devices. Not only that, Google launched Gemini for Gmail, which now means that over two billion inboxes worth of data is ripe and ready for them to train Gemini on into the most personal, best AI assistant you could have ever hoped for. But there's a third thing as well. Google also launched a product that might make them the backbone of the e-commerce industry. If your name is Jeff Bezos or Amazon, you're sweating buckets right now.

There was also one little thing. Oh, yeah, they became the second most valuable company in the world surpassing a $4 trillion market cap, beating Apple for the first time since 2019 and coming in second under Nvidia. Now, most people saw three separate headlines from Google this week. But on this episode, we're going to show you the master plan that connects all of them and why Google might have just pulled off the most important week in AI in 2026. Man, what a week. I mean, they didn't even ship features.

They just captured choke points. This was like a huge power consolidation week for Google. And it very much feels like they're assembling this full-stack mode. And I mean, I guess as a visual, you can imagine your iPhone's brain is Gemini. Your inbox is Gemini. Your shopping checkout is Gemini. It's not an assistant. They're creating this operating.

system for your life. And this started early on where I think the common joke and how I use my phone personally is I have an Apple iPhone for hardware, but I use all of the Google apps on my phone. And this is taking it to the next level where Apple has failed to generate artificial intelligence, Apple intelligence on device, and now they've had to defer to Google to do this for them. Is it a bad thing? No. In fact, I think everybody wins. But maybe we could start there and we'll get into what exactly is going on with this new Apple and Google AI deal. Yeah, I mean, before we do, I mean, it's, It's worth considering that Google three years ago was the laughing sock of the entire AI industry. They had a terrible model. Open Air was running laps around them.

And then in the last year, their market cap has gone up 65%. And I think you were telling me before we started recording. This is crazy. This is crazy. Wait so from April 2025 until today, Google's market cap went from $1.8 trillion to $4.8 trillion. In the span of like, what is that, nine months, 10 months, that's like unbelievable levels. of growth for the second most valuable company on Earth? I think it's because they have this very multi-pronged strategy, right?

Like the typical AI company is like, hey, we have this cool model. Isn't it awesome? Here's a chat bot for you. Google has that, and they have the infrastructure via TPUs in cloud, and they have the distribution via Gmail, Google Maps, and pretty much any kind of internet app you use today. They're just kind of like this all-in-one unicorn behemoth, and it's showing, right?

So let's get into the Apple deal, Josh. Run me through the highlights. What's going on here? Yeah, so TLDR, Apple is changing its brain over from Siri to Gemini, except it is keeping it under the Siri branding. So what you're going to get is all the promises of Apple intelligence, hopefully, except instead of powered by Apple intelligence, it's actually under the hood powered by Google.

Now, this might come as a confusing headline to some people, because previously it was announced that Apple was working with OpenAI. In fact, if you asked Siri a difficult question, it would offload that question to chat GPT. That is no longer the case. Apple decided, wait a second, we actually don't really think Open AI is the best fit. We think Gemini is the best fit. And here we see the reasoning why. They said they've entered into a multi-year collaboration under which the next generation of Apple Foundation models would be based on Google's Gemini model.

After careful evaluation, Apple determined that Google's technology provides the most capable foundation for Apple Foundation models. So this is a big deal. One, because they're now paying Google a billion dollars a year for this right. But two, because now they actually have a smart intelligence that they can integrate into the OS in a way that everyone had always hoped for. Right. Like they pretty much failed version one of Apple intelligence. It was just a complete and total failure.

But now they have an opportunity to do it again. And they're doing so with the model that, I mean, we believe to be most capable from a company that we really appreciate. And they have a relationship that has been standing for a really long time. If you remember from what it's got to be close to 20 years ago, like over a decade ago, Apple started paying, or Google started paying Apple for the exclusive rights to have Google as the search engine on Safari. Well, now Apple's paying it forward.

And instead of $20 billion, they're paying $1 billion. So it's 5%. But this is a really harmonious deal. I think everybody wins in the situation. And as an Apple user, I'm excited. I've always used Google apps. And now I could use Google AI in addition to those apps. I think Google was the only option that Apple could really consider when it,

came to putting an AI model in their devices. Why? Google is the only company that can scale to the lengths that Apple has. OpenAI is still a small startup, Anthropic, still a small startup. Google has the entire full stack, right? They have cloud, they have TPUs, they have all the distribution, and they have all the money and resources to be able to kind of scale this out to the size that Apple wants it to be. The other important thing is the quote you mentioned, which is they believe Google has the most capable foundation. I think that's a signal and a hint that the way that Google's model is created, the fact that you can ingest video, audio, as well as text,

makes it this kind of like all-knowing model, which is way more useful for Apple who builds products and apps and software in a very similar way. The other thing that I think is important to point out here is there's this kind of like platform play that is interchangeable between Apple and Google. As you mentioned, they've had a very long relationship. And in fact, Google has been the one that typically pays Apple for the privilege of making Google search the default search engine in Safari, right? And they pay them, I think it's $20 billion a year. And that's existed for so long now, right? But now we see the first instance of it going the other way around where Apple is paying

Google for the privilege of having their model embedded in their phone. And they're going to use it to train a personalized version of Siri and their own Apple foundational models. And I can't help but think that there's a non-zero chance in the future that this trend flips and Apple ends up paying Google more. Do you think that world would ever happen where like, so then how did they get to 5% of that total number if the expectation is that it can actually grow to $20 billion? What does that look like if they're paying Google $20 billion? What is that for?

Okay. So my take on this is, If or when rather, they do end up paying Google more, Apple will be making way more on the top end. The reason being is Apple is a product consumer-oriented company, and they are amazing at building exactly that, the consumer user experience. They do this through devices, integrated software stacks. It is just the best available ecosystem that you can access. You and I have used Apple for God knows how long, right? I think they're going to continue to do this.

That's where their bread and butter is. but they need the brain to be able to pull this off. To kind of simplify this, I think if Google builds the best AI model, Apple will build the best apps for everyone in the world to use. And that typically ends up making the most money if you look at every single tech cycle to date. Do you have a different day? I feel like they would swap out before getting to the $20 billion mark.

I'm just struggling to understand how they would be paying such a huge premium relative to where they are today. Like, where is that difference made up? Well, the switching costs becomes way too high, right? Imagine if every personal version of Apple's AI model that you use or that any of their iPhone users use is trained on Gemini, I'm guessing some portion of that data is trained on Gemini's, or rather Google servers to some extent, right?

Apple can't hold all of that data unless you think they're going to expand their private cloud to, like, own that part. I don't see Google giving that part up, right? That's what is going to make their models more competent in the future. Yeah, it seems like the strategy is going to be this two-fold thing, where they have Google and the cloud, and that's their foundation model that handles the difficult tasks. But they also have these really precise and very effective hyper-local models

that run very efficiently and are very small to kind of handle the local on machine compute. And we've talked about this in a few episodes before, but it feels like this is still an underrated strategy for Apple where they've all floated this AI responsibility, I guess, to Google in a big way. But in the little way, there's still a huge opportunity on the local devices for Apple to run their own, completely totally private models that run on this secure enclave. And again, if you're a developer who's building for these platforms, that's free inference. So maybe Google will give you really great rates, and it'll give you really great results.

But there's still this huge opportunity to unlock for anyone who wants to run these local AI models on Apple devices. And I'm hopeful that the developer community is going to start to see this because the cost of inference on these devices is zero. And I mean, if Google does continue to ramp up the price, I assume it will be kind of in correspondence to the increased compute of these chips that can then run more and more capable models even on a local device. So maybe there's a world in which Google doesn't actually need to provide a ton of the intelligence, just the really hard stuff, and eventually Apple's local models will catch up. I don't know, but it's an exciting thing to say a new paradigm shift, that they're actually taking it seriously now because

Apple intelligence as serious as it felt was certainly not that. Listen, there's a perspective that would say Apple is actually the genius in the AI race here because they haven't spent tens of billions of dollars to try and highly likely fail at building an AI model. We've seen that with meta so far, right? So, you know, they've just waited for the cream of the crop to rise to the top and then they pick the best model and now they're going to go do what they do best. Yeah, I'll take it for, exactly. I'll take it for a billion dollars.

Also, you're still paying. me $19 billion net to have the default search thing, right? So there is a world where Apple licenses the model, builds the best AI agent, which supposedly is going to be the main interface that will replace the search engine and the internet browser itself, which would mean that they could end up winning overall. So I guess I'm bullish Apple, Josh, which is funny because on previous episodes, I just haven't been that. So yeah, I'm feeling optimistic about Apple's deal here. Bullish Apple, bullish Google? I mean, hey, Apple, doesn't have to worry about this data center craziness.

They're in a pretty good spot. One man isn't happy, though. One man isn't happy, Josh. Who's that? That is Elon Musk. Oh? He commented on this whole news, and he said, this seems like an unreasonable concentration of power for Google,

given that they also have Android and Chrome. All right. Well, Elon, unless you're going to go make an iPhone, I don't know what to tell you, dude. I think the consumers are happy about this one. Look at this tweet from Quinn. Nobody wants to license Grok. little bro. It's gone.

Just wait until 4.2 comes. We don't until graph 5. Anyways, that's the Apple and Google News. We have more Google news as it relates to Gmail. If you are not using Gmail, I don't even know what you're using, but Gmail has some good updates this week from Gemini. So, can you please share the cool new Gemini features that we'll be getting in our inboxes starting today, I think?

Yeah. So the highlight is Gemini 3 Pro is now available in your inbox in Gmail. It can do a bunch of cool things. It can get rid of all the garbage emails that you're never going to respond to. All those unreads, just get rid of them. And it can prioritize the ones that you need to respond to. But also, it can train on all of your emails and understand your voice and what you want to say in your responses. So it can start writing up responses for you and automating a bunch of work that you just don't want to do.

But the thing that I found the most interesting, Josh, has got nothing to do with Gmail, ironically. It's to do with Google Workspace. which it also has access to. So it's not just a inbox. It has access to your Google Drive, your Google Documents, your Google Sheets, which is just a plethora of data that no other company has access to,

which brings me to like one of the key takeaways here and the story that I think a lot of people missed with this headline is Google just unlocked the mother of all data modes. Think about it. Claude has access to your conversations. Open AI has access to your conversations. Google has access to 2 billion inboxes. It has access to all your Google Maps data.

It has access to your search engine history. It has access to Android. It has access to all of this data, which they can feed into their model and train a better model, which ends up becoming more addictive for a person to use, a better experience,

which then feeds into this kind of like whole data mode. And to your point, they own the monopoly on email. I think it's something like 30% of the email market share. And Josh, get this. 90% of US startups also use Google Drive and Gmail, which means that they get access to all this data as well. Now, I think it's important to state

that Google has said that they're not going to use the data to train their models, but it doesn't stop them from using trend analysis and pattern matching to train their models, which is where the gold mine is anyway. Yeah, so as you were saying this, I was looking at how many people on earth use the internet and the number's about $6 billion.

And Google says about $3 billion of those, 50% rely on Gmail, which is just this unbelievably large amount. So like you described, it's great for data. It's great for Google in terms of user modes and lock-in. But I think there's an argument to be made that's also great for the user too. I mean, in a way, email is the closest thing we have to a life database. Like normally, normal people, they kind of maintain everything through there. It's like, this is where your receipts live, where you buy something, this is where your meeting invites live. This is where the contracts that you sign live. This is where the intros to people who you meet live.

It's kind of the place where all of the receipts, all of the cataloging of your life lives. And applying Gemini on top of that becomes this really powerful thing where it builds this life operating system on receipts that you've collected since you've started this Gmail account. And one of the really cool things

that I've been playing around with recently is Claude Code this week. And I use this tool named Obsidian where I've just written notes into it for years and years and years. And for the first time ever, it's able to take that backlog of data and actually parse through and make sense of it

and connect things in ways that weren't intuitive. And this feels like the lazy man's version of that where even if you haven't been writing into this entry database for years, you've been collecting these kind of things along the way. You've been collecting receipts. You've been collecting meetings and you've been collecting whatever it is.

And now Gemini is able to parse through that, make sense of it, and build a more global model that understands you well. So in that sense alone, it's awesome. The other sense is actually for writing emails. Are we getting to a place now where AIs are just going to be emailing AI? Like, you email me and you're just like, oh, just tell Josh, this, that, and the third.

And then I have my AI reply with whatever it thinks that I'm going to know. And then suddenly your Gmail just becomes automated. So this can go a various amount of directions. But I think directionally it's great. This is like really exciting for anybody who's a user of Gmail. Your experience just got way better. Yeah, it's funny. Technology, which is supposed to connect humanity as a whole, is ending up becoming the disconnecting engine of the entire world.

It's just because we're lazy, man. It's great at it. And like, why would I go write a long email when it could do it better than me, faster than me, more effective than me, or convincing than me, it knows everything about me. It's like this, it's this unbelievable thing. It's weird. It's so true. You know, you mentioned you're using code code.

I've been using Claude Co-Work, which is basically Claude Code for all the non-technical stuff over the last like 24 hours. And, dude, I currently have Claude Co-Work doing a bunch of tasks that I have like put off for an entire year. And it's quick. We're going to be putting out an episode of this letter this week. Yes.

Oh, it's my favorite thing. I'm so stoked for that. It is so cool. But anyway, to kind of like wind up the Gmail point, I think it's a really good example of AI coming to the people and meeting the people weather. Right now, AI has existed as this chatbot interface. You need to kind of download it, pay a subscription.

And now it's kind of this useful tool that everyday people can use. It's not just you or I, Josh. It is my mom. It is my sister. It is pretty much anyone that has a Gmail account. You know, there's 3 billion users, which I think is cool. The second thing is this just kind of locks in people to using Gemini AI. We had a conversation before we recorded this where you kind of raised the point,

why wouldn't I just connect Claude Code to this and connect to it via API? And the simple answer to that is most people aren't going to go through setting up an API account, which is a separate thing to Claude's subscription, by the way, and then connecting it, adding the API key, if they just have Gemini baked into it, which is just a full stack experience. Google wins again. Yeah, the switching costs go up.

Why would I want to just install this other thing? It's already here, and it's good enough. It may not be better than this particular benchmark, but I don't care. It's already here. It's trained on my memory. I'm locked in. I'm not using anything else.

Yeah, and that's why distribution is so important. You just meet the people where they are, and you have 3 billion people waiting to receive whatever you give to them. And today, it's AI. Now, EJez, for our third and final topic, I want you to explain this because if I remember just a few weeks ago, I read a newsletter published by you that said Amazon is going to be one of the biggest winners of 2026.

And as I'm reading the news this week from Google, and particularly soon to our Pachai with his post, it's telling me, well, actually, Google is going to work to build a competitor to Amazon. And maybe I'm reading this right, maybe I'm not, but they're teaming up with companies like Walmart, Shopify, Etsy, Wayfair, Target. to create this universal commerce protocol. So maybe you could explain the bull and bear case for this new protocol, what it is, how it works.

What's going on here? Is this a real problem for Amazon? So let's get into this. Google released a product or protocol called the Universal Commerce Protocol. And it basically will allow you or anyone that uses Gemini to make purchases and shop via the Gemini chat interface.

And the way that they do this is they hook up APIs from various different shopping partners, like you mentioned, Shopify, Walmart, etc., connects to your wallet. And now there's this seamless experience where Gemini, via the context that you provided it through previous conversations, also by evaluating your shopping history, will suggest products as you're having conversations with it or useful tools that might help you with a task that you're trying to kind of resolve by talking to Gemini and say, hey, do you want me to buy this for you? I'll get it delivered to you and it'll arrive tomorrow. Now, we've seen various

instances of this appear. I think Codex, which is OpenAI's first version of an agent, offered something like this, but it just never manifested. This is the real deal. It's still in a beta testing period, which is why I can't really comment on it thoroughly, but

it's supposedly meant to do the thing that we've been promised for over a year now. So then the question becomes, if everyone is doing their shopping via Gemini or via an AI chatbot, what happens to Amazon. And that's a real problem, Josh, because what this ends up putting Google in the

position is, is the intense layer. What do I mean by that? What I mean is it'll know exactly what you want to buy, and it'll purchase it for you, and it'll know why you want to buy the thing, which is really important. Amazon, which currently has this ability, because you go on Amazon, you scroll, so it kind of looks at you use this habit, will get that removed completely. So they'll just end up becoming the warehouse and delivery fulfillment center. Now, that's not exactly a bad thing. And I would actually argue, Josh, that that has been Amazon's bread and butter, right? What's the biggest moneymaker Amazon? It's not their shopping or e-commerce platform. It's AWS. They are an infrastructure fulfillment center company. To your point, they move atoms more than any other company

in the world. Big time. I think they'll just continue doing this. And Google will just be the layer on top that routes arguably more purchases to Amazon if they fulfill, you know, their vision of being, you know, the biggest distributor. Yeah, it's a good take. It seems right. It seems like in a way Amazon is kind of like the mall and then Google wants to be the credit card network plus the front door to that mall. And what we discussed in the last section where Google has three billion people, they have

the people where they are. In the case of Amazon, you kind of have to seek that out. You go to Amazon to buy something. You go there with intention, but the intention is always fairly indebted. intentional. With Google, you're just there. You're searching for carry-on luggage. You're searching for trips. And then it could integrate these recommendations like we're sitting on screen with this example into the search results. So that point about intention is really important because, I mean, is this bad for Amazon? No, because Amazon still owns the fulfillment, prime, the last mile

dominance that they have. But yes, in the sense that Google does own that upstream intent. And the moment you decide to buy something, Google can be there to capture that. So it's interesting. We'll see how this plays out. I mean, Open AI earlier this year did something similar where they partnered with a lot of these companies and they integrated it into the search results, but I haven't actually experienced that. I'm not sure if it's working, like, I haven't

bought anything through ChatGBT, and yet they rolled this out months ago. So I think the actual execution on this really matters because on paper it seems great, but is this actually going to convert sales using these new customers? That remains to be determined. Like, have you have you seen any, have you booked anything

through ChatGPT? Post and No, I thought it was actually a terrible product, and I remember being super excited by it, but it just never took off. And maybe we'll see the same story here. The real litmus test will be if my girlfriend starts using it. I'll end up with a problem at that point, right? Two other important points here is Google was also kind of like the doormat to internet shopping, right? Like the search links, Josh, 10 blue links, but then you had to kind of like figure it out from there. you had to browse and find the right product. And now they're saying, hey, we got the product for you.

Do you want me to buy it and send it to you? So there's this interesting evolution where kind of unexpectedly, the monopolist is still trying to monopolize on the new frontier, which is like pretty cool to see. The other thing I just thought of, Josh, is we mentioned a stat earlier, which is 90% of US startups use Google Drive, Gmail, the entire suite. So that technically means they could sell their product through Gemini in this new commerce protocol, which would be insane, right?

And would cement Google's distribution at this shopping layer. It should just be insane overnight. So that's it. I mean, big week. In conclusion, I guess this is how you become unstoppable, right? You don't just build the brain. You buy the distribution. You embed it into the memory layer.

You standardize the checkout rails. You kind of like capture the whole thing. And it feels like if 2024 and 5 were the model wars, 2026 is becoming the, I guess maybe the default war. Like where do you go by default to engage with AI? And if that answer is Google products, it stands to be a very good year for them. And we're seeing that in the chat GPT and OpenAI,

a degration of weekly active users. I mean, the relative market share of Gemini of Claude is having an unbelievable month. They're starting to kind of chisel away. I think Gemini hit 21%, right? of AI traffic. Something like that. It's happening quick. It's happening quick.

And one thing that I haven't heard in a while is an update to that 800 million weekly active user number from OpenAI. And I really am wondering how they're planning to deal with this because now Claude is building, did you see Claude built the health app this week too?

We're going to have to talk about this in the roundup. There's so much stuff that happened this week. It's unbelievable. But there's been big changes here and lots of paradigm shifts happening. And in terms of Google, this is a huge win. the board. It feels like it's a huge win for everyone, for Google users, for Google customers,

for Google shareholders, and for Apple users as well. So big wins across the board, maybe not for Amazon, maybe not for anyone competing with these companies, but nonetheless, a really exciting week for Google and giving probably everyone listening to this at least something new to try out and play with. Yeah, I agree. I bought more Google stock this morning before recording this episode. But I know a lot of you listening to this, of hate Google. I might be wrong here, but if you do, let us know why. Let us know your thoughts on this about this. I'm down. We do it every day anyway. A lot of you like to argue and we're there for it. But if you have any kind of opposing views and opinions, we want to hear it because

like we want to be able to know whether we're kind of like on the right track or not. Josh and I take this stuff pretty seriously. And we see a huge amount of potential in Google and we kind of don't see how they can fail, which is either a great opinion or a really, really bad one. So let us know in the comments. If you aren't subscribed or you don't have notifications turned on, please do both of those things. It's around 80% of you. In 20, 26, we would like to get that down to what, Josh?

30%, 20%, 0%. Yeah, we got to hit 30K subs too. We're right there and we need like 500 more. So if you want to help with that cost, tell your friends, tell your friends, friends, hit the subscribe button. It goes a long way in helping us start off strong in this new year because by the end of this year, there's a lot, there's lots going to happen and it's going to be crazy, it's going to be chaotic. And you're going to want to be here for the rikes. We'll cover it three or four times a week every single week. So tell your friends that it comes hopefully

packaged in these 25 minute episodes. We did pretty good today, I think. We kept it moving today. So hopefully this was enough for you to squeeze it into whatever window that you're listening to these episodes in. We'll see you next time where we talk about Claude because it's freaking awesome. So you're going to want to stay tuned for that one. Really cool. paradigm shifting platform. But yeah, until then, hope you enjoyed this episode, and we will see you guys on the next one.

See you guys.