(Preview) Google Starts Dancing, The Winners and Losers of Gemini Week, OpenAI Has an Advertising Problem

Ben and Andrew begin with Gemini 3, what to make of its terrific benchmark results, and why TPUs provide Google a sustainable cost advantage in AI. From there: Google’s opportunity in the enterprise space, Apple’s white label deal, and questions about both OpenAI’s future growth and challenges that loom if ChatGPT can’t incorporate advertising. At the end: Thoughts on what Gemini means for Nvidia, Anthropic’s market in AI, why Amazon and OpenAI are losers of the Anthropic-Nvidia-Microsoft announcement, and a correction regarding Charter cellular coverage.

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Andrew Sharp?

  Hello and welcome to a free preview of Sharp Tech.

• 0:08Andrew Sharp?

  Hello and welcome back to another episode of Sharp Tech. I'm Andrew Sharp and on the other line, Ben Thompson. Ben, how you doing?

• 0:19Ben Thompson?

  You know, Andrew, there is.

• 0:22Andrew Sharp?

  Oh, boy.

• 0:23Ben Thompson?

  Take like an NBA, take like an NBA basketball player. There is sort of two curves in their career. The first is sort of athleticism. When you come into the league, you're probably, you might still have a little bit more of a peak, particularly if you need to put on weight. Giannis would be sort of a good example of this. Comes in very spindly, very thin, puts on a bunch of muscle. But over time, you're going to peak relatively early in athleticism. Then it's going to kind of go downhill. And yet we don't say a player's prime is when they're 23 or 24. We consider it to be usually when they're 27, 28, 29.

• 0:59Andrew Sharp?

  29.

• 1:00Ben Thompson?

  Cause the other curve is the increase in experience and understanding and how to actually play basketball.

• 1:06Andrew Sharp?

  The game slows down before you go in your career.

• 1:09Ben Thompson?

  It's good that the game slows down because your body slows down as well. And so there's, there's this, this inner, this, this period where the athleticism has declined, but not enough. And the experience has increased to more than make up for it. Eventually, however, you can be as experienced as you want and you just, you don't have it athletically anymore.

• 1:30Andrew Sharp?

  The lines on the graph diverge a little bit too much. Yeah, Blake Griffin, actually on Amazon, their new NBA coverage, Blake Griffin was talking about the moment he knew he had to retire. He was running up and down the court with Evan Mobley and he was playing like three or four minutes. This is Blake Griffin on the Boston Celtics, like three or four years ago. And there was a foul and he was standing at the free throw line next to Evan Mobley. And he looked up at Evan Mobley and Evan Mobley had been running up and down the court like nothing happened. And Blake Griffin has his hands on his knees and is exhausted and is like, man, it's time to go. So. Is that what you're confronting? That sort of moment right now?

• 2:12Ben Thompson?

  The experience is you just learn how to do things, right? You learn how to write, turn something out every day, to be a professional podcaster. Even when you're not feeling it, you're feeling tired, you sort of get amped up, you're ready to do it. You learn how to do things like teas. Like there's segments in this podcast that I'm going to give a little hint towards. They involve getting old.

• 2:34Andrew Sharp?

  Okay.

• 2:34Ben Thompson?

  And, you know, so I like to think that my experience is improving and is enough to compensate for the horrific reality that we have a question on this rundown that has been on the rundown for a couple of weeks. I was sure we had addressed this question previously. I got in an argument with you. I brought in Dumbin to review, professional podcast listener Dumbin. He's like, no, I don't think you talked about that. And you know what? My memory on which I've depended to a near absolute basis, my entire life, just sort of unknowingly, there's so many things that are based on, I just remember, be like, how do you cite that article from 2017? I'm like, well, I just, I don't know, I just thought of it and decided to link it.

• 3:20Andrew Sharp?

  Yeah.

• 3:20Ben Thompson?

  Fading, fading fast, Andrew. I am facing my mortality. I am praying the experience. I'm hoping that AI can save me.

• 3:28Andrew Sharp?

  Well, at least yours is a colorful age-induced dementia because you were effectively imagining podcast conversations that we did not have on the show.

• 3:39Ben Thompson?

  It's true. So like, so I need the readers, if there is a question on this podcast that you think we covered before, you have to email in.

• 3:47Andrew Sharp?

  Email in for Ben's sanity. Exactly. Make Ben feel better over the Thanksgiving holiday there.

• 3:52Ben Thompson?

  As long as we're, we're, we're sort of looking ahead. I have, I have, there's a topic on here that is going to look ahead to Thanksgiving.

• 3:59Andrew Sharp?

  Another tease, professional podcaster.

• 4:02Ben Thompson?

  That's right. And we will do a Thanksgiving episode next week. Per tradition, unless there's big news.

• 4:09Andrew Sharp?

  Yeah.

• 4:10Ben Thompson?

  It was Thanksgiving three years ago where the OpenAI stuff went down.

• 4:14Andrew Sharp?

  That's right.

• 4:15Ben Thompson?

  We know, because it was the Las Vegas GP. It was the first one.

• 4:18Andrew Sharp?

  And we had to keep delaying the podcast. We were gonna record in person in Las Vegas and we had to keep delaying it because the news kept coming hot and heavy that way.

• 4:27Ben Thompson?

  And I was, and I got back Sunday night and I was delaying writing, which was good because Satya Nadella drops this tweet at like 2:00 AM like, oh yeah, we're acquiring whatever my Mariah Sam Altman did wrong with. Uh, man, what a weekend that was. So barring something like that, we are going to have our silly, off-topic, massive mailbag. If it's about tech, it has to be clever and funny. We're not doing like strategy breakdowns, X, Y, Z. We're gonna have a fun time.

• 4:55Andrew Sharp?

  Indeed. And if you've got questions for that episode, please submit them by 8:00 PM Eastern on Monday. Ben and I will record the episode on Tuesday afternoon, but I gotta put the rundowns together, the rundown together Monday night. And by the way, as bad as you feel for imagining podcast conversations that we didn't have, I am heartened to know that you actually read the rundowns before we come on to record every episode. This is evidence.

• 5:24Ben Thompson?

  We did have this knockdown, throwdown fight approximately eight minutes ago, so don't give me too much credit, but yes. No, actually, you know what my thought was? My thought was I should not read the rundowns because if I hadn't read this rundown previously and seen this question, which has been on there for a couple of weeks, I wouldn't have had this problem. So actually, I'm being punished for being prepared.

• 5:43Andrew Sharp?

  Well, in any event, we are going to begin with Gemini and three quotes from 2023. Sam Altman in February 2023 said, "If I were sitting on a lethargic search monopoly and had to think about a world where there was going to be a real challenge to the way that monetization of this works, and maybe even a temporary downward pressure, I would not feel great about that." And then Satya Nadella.

• 6:11Ben Thompson?

  Where did he say that?

• 6:11Andrew Sharp?

  What's that?

• 6:12Ben Thompson?

  Where did he say that?

• 6:13Andrew Sharp?

  He said that on Strategery, in a Strategery interview. That's right. I should credit the outlet here, okay? Sam Altman to Strategery in February 2023. Satya Nadella to The Verge in February 2023.

• 6:29Ben Thompson?

  We don't need to cite everyone.

• 6:31Andrew Sharp?

  At the end of the day, Satya said, they're the 800 pound gorilla in this. That is what they are. And I hope that with our innovation, they will want to come out and show that they can dance. And I want people to know that we made them dance.

• 6:48Ben Thompson?

  Look, I keep using the Altman quote because it's honest to checkery. But this is, this is the killer quote.

• 6:53Andrew Sharp?

  It's terrific.

• 6:55Ben Thompson?

  It's so good.

• 6:56Andrew Sharp?

  And then another quote that I had missed at the time, but I saw as I was prepping. Google CEO Sundar Pichai with a nice mild mannered response. One of the ways you can do the wrong thing is by listening to noise out there and playing to someone else's dance music. Totally missed that at the time, but a wonderfully nerdy response from him. So there you have it, Ben. That's where we were two years ago. Google moving to its own beat. Two years later, Gemini is here. It's beating the competition on every model benchmark but one. And we'll talk about the implications for a variety of AI players at the top here. One threshold question, though.

• 7:41Andrew Sharp?

  As an analyst, how do you handle the release of a new model like this? Like, how much stock do you put in various benchmarking tests?

• 7:51Ben Thompson?

  It's really tough. And it's actually something that I've thought a fair bit amount. And this is sort of going to be an issue, challenge, question, and reason for optimism for OpenAI, actually, which I think we're going to get to in a bit. Perhaps per our recycled question.

• 8:09Andrew Sharp?

  There you go. Another tease.

• 8:13Ben Thompson?

  The reality is, is AI is already so good for so many situations for so many people. It's harder and harder to perceive what differences there are.

• 8:21Andrew Sharp?

  Yeah.

• 8:22Ben Thompson?

  Now, you could argue that that's because you're too stupid, which very well may be the case. At some point, AI is like.

• 8:29Ben Thompson?

  It's easier, how to put this? If you're smarter than different people, you can perceive the differences in intelligence. If you're not as smart as other people, they're all just smarter than you. And one of the fortunes of, you know, being in Silicon Valley and getting to know lots of people. I've definitely, I've talked to people that are definitely much smarter than me. And I'm not sure how much smarter than they are. They're just smarter than me. And I've talked to other people and I'm like, I can kind of rate you guys. I don't know who's where. So, and it's, you know, it's sort of an uncomfortable topic, but you just have, you can see down easier than you can see up.

• 9:08Andrew Sharp?

  Yeah.

• 9:09Ben Thompson?

  And one of the challenges with AI is if it's not there yet, it's going to get to the point where it's sort of looking up for everyone. And so like, what's the difference? Sort of X, Y, Z. Now, AI also tends to be very spiky. There's some stuff it's just really good at and some stuff that you can kind of tell it's kind of making it up. And it will hallucinate much more in particular topic areas, or it will just sort of get stuff wrong. Or you could really see the seams of, you're just pulling straight from Reddit. And I'm not sure that, you know, it's sort of aspects to it. But probably the most useful, important area for the benchmarking are areas like coding, for example. And the reason is they're verifiable.

• 9:53Ben Thompson?

  And this is a point that I've been making again and again. Karpathy sort of made out a tweet along similar lines this week, where AI is the most interesting at the two ends of the extreme. The one extreme is where there's no right answer. So you're like, this is where hallucination is actually a positive trait. It makes up new stuff. That's actually incredible. Computers have never done that before. The other extreme is where you could, you know if something is right or wrong. This is why code, I think, has been a huge area because the code has to run.

• 10:28Andrew Sharp?

  Mm hmm.

• 10:29Ben Thompson?

  Like, like it either runs or it doesn't. It compiles or it doesn't. It throws errors or it doesn't. And in areas that are verifiable, not only can you know how good it is, but also you can do reinforcement learning where you can do post-training to sort of make the base model even better by running it through a bunch of tasks and then learning from them and getting sort of better over time. So what's interesting about these benchmarks is, by the way, the one benchmark Gemini did not come in first in was coding.

• 11:01Andrew Sharp?

  Mm hmm.

• 11:01Ben Thompson?

  Anthropic is still number one. Anthropic figured something out with, I think it was 3.5, that was like blew everyone's mind as far as coding and it's been a sustainable, persistent lead, which I think is probably surprised, it surprised me probably a bit. I don't know what, what secret sauce there is and if, is, is it going to be figured out sort of eventually?

• 11:22Andrew Sharp?

  Right.

• 11:22Ben Thompson?

  But that's a very positive sign. Yeah, like I think Anthropic was actually one of the big winners of this week, where their bread and butter is the one thing that Gemini did not beat, at least according to benchmarks.

• 11:33Andrew Sharp?

  Yeah, they have them for now.

• 11:35Ben Thompson?

  Well, the other problem is the benchmarks are public. Now, they do their best to keep the questions secret and X, Y, Z and, and adjust them.

• 11:43Andrew Sharp?

  That's part of what I wonder about is like, are models just being trained to the benchmarks at this point? And how much do we really learn from?

• 11:51Ben Thompson?

  Well, sometimes explicitly, like Lama 4 released like this benchmark specific version and it scored super high. And that was a real, like, I don't understand the people who are critical of people in meta AI being, set aside the super intelligence and who they hired.

• 12:08Andrew Sharp?

  Mm hmm.

• 12:09Ben Thompson?

  If you have a team that is disguising their failure by cheating.

• 12:16Ben Thompson?

  Heads need to roll. Like, there's, there's multiple issues going on, not just in the failure, but in the cover up and all those sorts of things. And, you know, Mark Zuckerberg, one of the many reasons he looked bizarre, I think, in the interview I did with him last year, and kind of dumb in the, I think, the interview with Dorkish in the same week, is he's sitting out there on a spit, trying to defend the fact that he has a team that was cheating on benchmarks to cover up their failure.

• 12:43Andrew Sharp?

  Yeah.

• 12:43Ben Thompson?

  Like, like, that's, that's just not, it's not okay.

• 12:46Andrew Sharp?

  Their failure being effectively a mediocre model.

• 12:49Ben Thompson?

  That long form wasn't very good. Yeah, that, that it was, it was a sort of, sort of a poor model and they'd sort of gone the right direction with this monolithic sort of approach. And now, the funny thing about Google and Gemini, this is not an assertion about Gemini 3. All I will say is that there was a friend I was having a discussion with, and this was a while ago, I want to say like six months or so, where they were sort of tongue in cheek, characterizing the various AI teams. And the, the line about Google is their, whether they're weak, there's something about the weaknesses in, in, in the teams. And I think something about Anthropic was sort of the, the goody goody two shoes bit and like, oh, we're so special, blah, blah, blah. The OpenAI bit, or they said that they reflect their founders, I can't remember exactly what it was.

• 13:34Ben Thompson?

  I'm so sorry, I don't want to speak for this person who I'm not saying because I'm butchering whatever they said.

• 13:39Andrew Sharp?

  I was gonna say, I have no idea what was trying to be said.

• 13:42Ben Thompson?

  Right, the OpenAI one was, uh, uh, you know, the complaint about OpenAI is it's very, uh, it's sort of like kisses up to you all the time. And it's like just telling you what you want to hear. Right, and saying like, yeah, that's Sam Altman as well. Like, that was sort of the funny one as well. But Google was, uh, yeah, Google builds to the benchmark. That's part of the beer crack nature of a large company, you get what you measure.

• 14:04Andrew Sharp?

  Sure.

• 14:04Ben Thompson?

  And if that's what being built towards, right. And so it's just interesting that they made this observation to me a while ago, and here we are today and Gemini is blowing the doors off in terms of benchmarks. But does that actually translate?

• 14:19Ben Thompson?

  In the real world, I don't know. It's really hard. So from my perspective, what I do want to do is use Gemini regularly. Like when Grok four came out, it was a big lead forward and I used it regularly. And then I got sick of using it in the browser on my desktop where I use it the most. And it got a little worse. Chat GPT seemed to get a little better. I went back to GPT. So, you know, so I do, I do need to use it more. What's interesting, for what it's worth, and this is pure anecdotal, which means it's probably not worth anything. There were two extended discussions I was having with ChatGPT. I started on ChatGPT.

• 15:00Ben Thompson?

  One was about a networking issue and one was about turkey preparation. Now, the networking issue, I know enough to be dangerous, but I wouldn't say I'm an expert. Uh, what I will say is that the ChatGPT answer blew the Gemini answer out of the water. It actually answered the question. It gave specific references. And it said why the answer to the solution that I wanted was probably not a good idea to do. So I have all ubiquity unify equipment. I wanted to see, can I limit a, like the guest network, the bandwidth and you can do this through quality of service policies. But that's going to kill your overall networking because, because it's no longer on a dedicated, like hardware channel, it does doing software, your whole thing will be slowed down.

• 15:44Ben Thompson?

  So it gives me this whole layout. Yes, you can do it. This is how, but you might not want to do it because of why. Gemini didn't even list the QOS option. Their option was get completely separate access points for your guest network and put it on a hundred megabit router or switch.

• 16:00Ben Thompson?

  Number one, bizarre answer. Number two, there was an answer. It didn't get, so anyhow, but whatever that happens. Sure. Yeah, that happens. I don't want to give one example. Now, what I am an expert at, I would like to think, is turkey preparation.

• 16:12Andrew Sharp?

  Great.

• 16:14Ben Thompson?

  And, uh, you know, I am a big believer in the wet brine approach for turkey preparation. Uh, the problem with turkey, people complain that it, that it's dry. Uh, you, I have my own brine, uh, for poultry that I think is phenomenal. My chicken, world-class, and I've done turkeys in lots of different ways, done in the oven. I've actually done on the grill, worked very, very well. I'm gonna do a smoker this year.

• 16:36Andrew Sharp?

  Have you fried a turkey yet?

• 16:38Ben Thompson?

  I have not. I don't have a big, I might try that in the future. I'm gonna try the smoker this year. That's sort of, sort of a new thing. But I know, I, I do have the brining process down pretty pat. And I was sort of reviewing, like, what's the timeline from thawing the turkey to do?

• 16:50Andrew Sharp?

  Prepping. Love it. Leaving no stone unturned before next week.

• 16:54Ben Thompson?

  No, because you gotta start early. You, it takes three, three or four days to sort of thaw, and then you would brine, and then the key thing is after the brine, you need to leave it uncovered in the refrigerator so the skin dries out, so you'll get crispy skin.

• 17:08Andrew Sharp?

  Mm.

• 17:09Ben Thompson?

  Which ChatGPT covered all that, gave me a good timeline. I'm like, oh, I should put this in Gemini. Gemini skipped the drying out the skin step.

• 17:17Andrew Sharp?

  Oh, boy. Gemini forgot to throw it in the fridge to let the skin dry.

• 17:22Ben Thompson?

  Really? I mean, you want to talk about a confidence shaker. We're talking about Thanksgiving dinner. And, uh, because it started out, it's like, oh, start thawing on Sunday night. I'm like, what? That's way too late. What are you talking about? I go through, it's like, I'm gonna pull it out of the brine and throw it straight on this. No, they're gonna get soggy skin. It's, oh.

• 17:41Andrew Sharp?

  Oh, boy. Wow.

• 17:42Ben Thompson?

  Anyhow, uh, nothing to do with probably the economic implications of, uh, AI, probably totally irrelevant and just bad luck for Gemini, but it speaks to why this is very difficult. Like.

• 17:54Andrew Sharp?

  It's mystifying. For me as a normie, it's mystifying to know how much stock to put in any of these tests. And every model release is hyped for a week or so in advance. And John Coogan of TBPN, he actually, he compared these releases to iPhone releases where, like, 10 years ago, there would be mind-blowing leaps that would make the whole world stop. And now the improvements are steadier. And less sexy than they were a year and a half ago, which isn't to say they don't matter. But vibes wise, it's just sort of a shift as the ecosystem grows up here. Does that resonate with you at all?

• 18:34Ben Thompson?

  Well, I think so, and it speaks to a couple of factors here. Number one, Google does have the comeback narrative behind them. And like narratives matter, right? You, that's how you frame this discussion from the beginning. And as a, you know, longtime professional NBA podcaster, who is one of the biggest proponents of narratives, led to one of your really terrible takes that Russell Westbrook's of an MVP as a six seed a few years ago, just because you love the narrative. This is a great narrative, right?

• 19:02Andrew Sharp?

  Yeah.

• 19:03Ben Thompson?

  There's also sort of systemic, structural reasons why you would expect actually Google to have the best model. They are of the most resources. They've been working on it the longest. They have this fully integrated stack, which at the beginning, you would expect the integrated player to actually do better. The whole theory is by integration with stuff's not good enough, you can solve a bunch of small problems and do it better. Over time, as you get the iPhone level where everyone finds it boring, it starts making more sense to have component pieces and people competing at different levels. We, we had the opposite, like we, Google invented the transformer, but then we had OpenAI running on Azure using Nvidia, like a much more modular approach,

• 19:45Ben Thompson?

  coming in and being in the lead. And so there's a bit where my,

• 19:53Ben Thompson?

  even with all my Turkey problems and my networking issues, my base assumption is that the benchmarks are broadly correct. Because I would expect Google to eventually get the best model. And if it's now, or it's the next one, whatever, I'm not surprised.

• 20:10Andrew Sharp?

  Interesting. Well, I mean, we're going to go company by company. You wrote about winners and losers on Wednesday after Gemini week here. But one specific question I had on Google and the integration, Google trained this model on TPUs, their tensor processing units, not Nvidia GPUs. So for Google's prospects specifically, how significant is that detail?

• 20:34Ben Thompson?

  Pretty significant. I mean, it, it, I think everyone had this sense that TPUs, you know, which they started a decade ago, and particularly the latest versions, like, like Nvidia GPUs, actually, to be fair, were not.

• 20:48Ben Thompson?

  Tailor-built for sort of LLM, sort of workloads. And that's actually more important for the TPUs because they're less capable, less programmable than a GPU. It's like a simplified set of what a GPU can do, but a GPU actually does more stuff and it's more programmable, can do more things. Part of the reason why they're more expensive and more complicated, not just expensive because Nvidia charges a lot, more expensive to make. Like, and more complex chip, larger chip, means you have more yield issues, et cetera, et cetera.

• 21:15Andrew Sharp?

  Right. Theoretically, they should be able to serve AI at a lower number than everybody else with TPUs.

• 21:22Ben Thompson?

  Yeah, especially once they get the right TPUs in place, in theory. And then, which those TPUs can be developed, again, at least in theory, hand in hand with their model development team. So the model, as they develop, decide how to train the model, for example. Like I mentioned, Llama was monolithic. It was one big clump of digital neurons, sort of figuring it out, as opposed to a mixture of experts' approach, where you have like different ones that light up in different places, which presents its own problems as far as cross-talk and traffic and communicating and making sure you're not overweighting one, et cetera, et cetera. But, so there's real, and this was in the context of DeepSeek, was a very aggressive mixture of experts' approach.

• 22:00Ben Thompson?

  And they did it in a sort of unique way that let them do it with relatively low capabilities, which we should go back to in a moment. But, so Google, there are architectural choices that go into training. And Google can do that with the knowledge of what TPUs are capable of. And they've been really focused with the last generation, I believe, of TPUs, on having much more scalability in terms of, you know, not just one big, but they could actually be much more stronger assumptions around being separate. And even potentially in separate facilities over time. And they're sort of building towards that, so you don't need one monolithic huge data center. You could actually do something that is in different locations, for example. All of which introduces a ton of cost and complexity, X, Y, Z.

• 22:41Ben Thompson?

  The more you can build with those in mind, at least in theory, the better. So this is a validation, not necessarily of the different locations thing, but whatever they're doing between the model building and the GPU design is paying off. And the implications are... They have a sustainable cost advantage going forward.

• 23:01Andrew Sharp?

  Mm hmm.

• 23:01Ben Thompson?

  Just number one, they're not paying Nvidia margins. Number two, TPUs are cheaper to make than GPUs. The margin one is probably the more important one. They get to internalize all that. Yes, Broadcom helps in some respects, but they're doing most of the secret sauce. Broadcom's margins are going to be anywhere near what Nvidia is because they're not making the chip. Google is using some baseline. Like it's more they they get a percentage of stuff, not they get to charge whatever margin they want on top of it.

• 23:26Andrew Sharp?

  Mm hmm.

• 23:27Ben Thompson?

  So, yeah, so you have, and this is the real sort of, the economic threat here is substantial. Google has huge cash flow, uh, better than ever, by the way, uh, you know.

• 23:41Andrew Sharp?

  More cash than anybody at potentially cheaper costs than everybody else.

• 23:46Ben Thompson?

  That's right.

• 23:47Andrew Sharp?

  Yeah.

• 23:47Ben Thompson?

  That's right. Uh, pretty devastating combination. And so to the extent that is validated is a big deal.

• 23:54Andrew Sharp?

  Mm hmm.

• 23:55Ben Thompson?

  We can get to what that means for the other companies. Uh, but yeah, it it's validation. It's not a surprise, but it is important validation.

• 24:02Andrew Sharp?

  Yeah, so if Google's model Gemini three is in fact best in class, what are the opportunities?

• 24:09Ben Thompson?

  Let's, and I do this in my daily update. Let's assume it is best in class.

• 24:13Andrew Sharp?

  Yeah, for the purposes of this conversation.

• 24:15Ben Thompson?

  Qualified to say that for sure.

• 24:17Andrew Sharp?

  That they killed it here. What opportunities for Google interest you most going forward in light of this success?

• 24:25Ben Thompson?

  Well, what's interesting is at least in in past era, in the smartphone era is probably the best example. To go back to sort of the beginning of stratechery. I've told this story a million times, but at the time when I started, the broad sentiment was that's neat that Apple invented the iPhone, but Android is going to obliterate them because modular beats integrated, blah, blah, blah. Look what happens to Windows versus the Mac. Um, as I've noted, the Windows versus the Mac analogy is actually widely misunderstood, most prominently by the late Dr. Christensen, uh, which is like, just totally wrong. Like, it wasn't the case that the Mac was first and then Windows came in and ate their lunch.

• 25:09Ben Thompson?

  It was the case that DOS was first, riding on the coattails of IBM, built up a huge customer base and software ecosystem. Then the Mac launched, and yes, the Mac was the first GUI, but Windows, which came later, ran all the DOS applications. So actually, Windows was first when it came to the install base and all those sorts of things. It's not an applicable answer. And the iPhone was different because the iPhone actually was first. And they built up a customer base and they built up a developer base that was not going to just leave and go to Android. It was like, it's, it's funny, like, just how these myths about the past permeate basically,

• 25:50Ben Thompson?

  it's like, everyone, and everyone's analysis is totally wrong because of it. I mean, like, it's like the lowest-hanging fruit stratech we ever had. Because like, if you actually change one of these facts about your understanding of what happened 20 years ago, your conclusions are gonna be totally different. And boy, I could build a business just staying this very obvious thing.

• 26:10Andrew Sharp?

  And now here we are. No, that element, I was not aware of that history until about a year and a half of hosting this podcast when it came up. And I cannot believe that, basically, people just got the facts wrong in terms of the sequencing of what happened and that informed.

• 26:30Ben Thompson?

  Well, because the sequence was that the Mac came before Windows, and Mac famously sued Microsoft for stealing the concept, which, of course, Apple stole from Xerox or borrowed, a great artist, whatever it is. So it's understandable that that is in the popular conception because the Mac did have the first GUI, Limp Interface. Microsoft was second, but it misses the actual important layer was the software layer, third-party developers in particular, the underlying OEM ecosystem, all that came before the Mac.

• 26:58Andrew Sharp?

  Okay, so how does that relate to where we are and where Google might go?

• 27:02Ben Thompson?

  Actually, it's interesting to contrast these two. And now, I'm actually regretting my update yesterday, I didn't dive into this. Let's get it, so Sharp Tech is for.

• 27:11Ben Thompson?

  So. If you look at the smartphone era, that's sort of what you would expect where, or at least what I would expect. So the big article I wrote about this, it was like the article, arguably the article I started struggling to write was why Clayton Christian was wrong, which is like, you know, it's a classic of the genre. You come up respecting someone, learn from someone, they're your hero, but to truly sort of like break through, you gotta take them down.

• 27:37Andrew Sharp?

  Take a shot at a legend, make a name for yourself.

• 27:38Ben Thompson?

  That's right.

• 27:39Andrew Sharp?

  Of course, yeah.

• 27:40Ben Thompson?

  And so what he was wrong about was the insistence that modularization would always be integration in the long run. And my big, one of my takeaways in that article is actually consumer markets are different because they're, the user is the buyer. There is no ceiling to the quality of the user experience. This sort of led to, you know, Eugene Way really expanded on this with the sort of invisible asymptote idea of, you know, you could just always get better and better and better, better in this respect. And so that meant that, that actually no, Apple wasn't going to be disrupted because consumers were not just going to decide, oh, this very abstract sort of spreadsheet driven calculation of like value relative to capabilities and suppliers and all sort of thing. Oh, I'm going to switch to end right now.

• 28:26Ben Thompson?

  All the examples Clayton Christian used were all in B to B businesses where you have actual, people are paid to be rational, to do sort of spreadsheets, to decide which one to choose X, Y, Z. Consumers aren't like that. And I pointed to, look, all these examples, I think it was like autos or consoles or lots of other things. Console was definitely a big one that stayed integrated because their focus was on elevating the user experience in a market where the buyer is the user. And so they're going to value that and design matters in the consumer space in a way it doesn't always matter in the enterprise space. If you're the spreadsheet guy making buying decisions, yeah, one of your columns in your spreadsheet is user experience. It's not the driving factor.

• 29:11Ben Thompson?

  When you have to use it, that's why people complain about their enterprise software, right? Because it's not, there's lots of other factors.

• 29:17Andrew Sharp?

  It clearly has not been the driving factor for white collar work over the last 30 years.

• 29:21Ben Thompson?

  Right. Well, I mean. To some extent, it did get better. And this was, this was a great thing about the SaaS era. And the whole sort of bottoms up philosophy of you, you win a team and then it spreads throughout the company. The good thing about that is that necessitates it being good. So you get, you know, evangelists within your company that buy stuff and sort of spread it around. And so the quality of enterprise software went up dramatically in the SaaS era relative to what was before. But still, nothing like the consumer market and.

• 29:49Andrew Sharp?

  I'm thinking of the battles I waged with Microsoft Teams on a weekly basis. That's what comes to mind.

• 29:55Ben Thompson?

  Still pretty rough. There's a reason we do use Teams for various things at Stratechery and all our conversations happen in WhatsApp because it is much better. So, uh, the, so, so generally, we saw the smartphone era and I think by and large, integration is particularly powerful in the consumer market. Because you can deliver a better user experience if you're just managing the whole thing from month to end. Even if you're giving out certain efficiencies along the way, those efficiencies, having modular interfaces, are costly in terms of the user experience because you're, there's just rough edges that you're having to tie together. And as a rule, I think that's a useful thing to think about. The problem is that we're not starting from scratch today.

• 30:40Ben Thompson?

  ChatGPT is in the markets. They have 800, 900 million weekly active users. Now we have weekly active users. It's a monthly or daily, whatever.

• 30:48Andrew Sharp?

  Yep. 800 is the number that was reported recently.

• 30:51Ben Thompson?

  So it's a much more difficult task to get people to change than it is to win a market fresh. So even if Gemini is better, how many people are actually going to even try it? Given that ChatGPT is still pretty good and better for unified networking and cooking turkeys.

• 31:11Andrew Sharp?

  Totally. I mean, that's what I wonder about as everybody freaks out about Google. I mean, it seems like there's been an awakening of sorts to Google's dominance, but I'm unclear how much Gemini's model performance really moves the needle with that story. Like the big win here. Is preserving search dominance over the last couple of years and growing the cloud business and YouTube, et cetera. Like, I understand why there's been an awakening to Google's dominance. It's just unclear what Gemini really means in that story because everyone I know still uses ChatGPT and not Gemini. And maybe that'll change, but I would guess not in the near term.

• 31:49Ben Thompson?

  Well, this is where it's interesting to go back to the PC era and the whole point I just made that the PC era started out modular.

• 31:56Andrew Sharp?

  Hmm.

• 31:57Ben Thompson?

  And then the integrated player came in. It got a rabid fan base that was also very small and never really got large.

• 32:07Andrew Sharp?

  Yeah.

• 32:07Ben Thompson?

  And that might be the better analogy for the consumer market in AI, which is the Mac was better, but it didn't matter because everyone was already using Windows.

• 32:21Andrew Sharp?

  All right. And that is the end of the free preview. If you'd like to hear more from Ben and I, there are links to subscribe in the show notes, or you can also go to sharptech.fm. Either option will get you access to a personalized feed that has all the shows we do every week, plus lots more great content from Stratechery and the Stratechery Plus bundle. Check it out. And if you've got feedback, please email us at [email protected].