Is This The End For Amazon, TSMC and META?

1. 0:00 – 0:40

   Intro

   1. BTBen Thompson0:00

      In retrospect, that was the seed of Amazon's kind of disastrous last few years. [laughing] They have an absolute absurd number of features, and the interface is pretty terrible. The metaverse stuff, I think it's a bad idea. Everything about their strategy works against that.

   2. BGBen Gilbert0:15

      Ben, I have one more section that I wanna do, which is, you do all this analysis on companies, and you have this very enviable position of getting to comment on and critique their strategies after they announce them, or an event after they do them. And I wanna play a game where I give you a company, and I'm curious, either if you were the CEO, or you were giving advice to the CEO, what would you do strategically? And I think an interesting place to start, 'cause we talked about them a bunch on this episode, is Meta.

2. 0:40 – 4:08

   Meta successes and failures

   1. BGBen Gilbert0:40

      You are the new CEO of Meta. What do you do over the next five years?

   2. BTBen Thompson0:45

      Well, I think Meta is actually doing, um, a fair number of things that they should. Now, the Apple changes were very devastating, and they're structurally devastating, and so they deserved a significant haircut on their valuation because of that. But I also think they were moat-enhancing in the long run, where I, I, I question whether any other company's ever gonna build a top-of-the-funnel advertising product that will be competitive with, with, with Meta, assuming they can keep their audience. It appears that they've done a good job limiting the sort of the TikTok threat. TikTok's growth has sort of flattened out over the last few years. So I think they're actually in good shape. What I do worry about is this is a company that's always been growing. It's really hard for those companies to shift to more of a scarcity mindset. I mean, they laid off 11,000 people, which only brings them back, like, nine months. They probably need to be a lot smaller, but that's also very destructive to company culture and morale. Do they still have the people that they need to pull that off? Those are probably some of the bigger questions, and obviously, any aggregator is dependent on having that audience. But I do think that the network effects of their products are still underrated. I mean, everyone has it in their head that they're shrinking, and actually, every product they have is still growing, which I think is underappreciated. So number one, I think they're kind of broadly on the right track. Number two, I would acquire Shopify, and I would take the FTC and Justice Department to court when they sue to stop it. I think they need to close the loop on e-commerce and advertising, and, well, again, I think that'd probably be bad generally, but I think it'd be very good for Meta, so I would do that. As far as the metaverse stuff, I think it's a bad idea. It's very hard to talk about the metaverse, 'cause it, like, what are the prospects versus is it, you know, X, Y, Z? I do think that it's not just a bad idea because it's taking so many resources and attention, I think, from Mark Zuckerberg, but also, I just don't think innovation is necessarily born of mass expenditures in large companies. We sort of skipped over a period of experimentation and ecosystem building that, in the long run, perhaps would be consolidated into a couple companies, but instead, it's just one monolith sort of trying to brute force this sort of bit. And the reality is, is Facebook is a services company. It's a social network, and everything about their strategy works against that. To succeed, they not only need to sell headsets, they need to sell headsets at sufficient scale and into friend networks such that people can social network on the headsets. And, and so their, their place of winning is, is even further away than I think people think, and so I don't think strategically it's the best thing for them to be doing, and I've been pretty anti-them doing from day one. When they bought Oculus, I said it was bad. And again, it's very hard to distinguish between, well, what are the prospects of this versus zooming out, and should they even be doing this? But I think the market is overly down on Meta, in part because the [chuckles] branding was too successful. This is still a powerhouse in social media and advertising. Uh, and the amount of money they're spending on the metaverse is, all things considered, not that much. They still, you know, have a $5 billion profit a quarter. So yeah, I would double down on what they are, and I would consolidate, not acquire Shopify, and spend three years fighting it out in court, because I think the payoff would be worth it.

   3. BGBen Gilbert3:59

      All right, the second one, rather than going with another big American tech company, I wanna go with one that's an Acquired fan favorite, and one that's very close to home for

3. 4:08 – 7:18

   TSMC analysis

   1. BGBen Gilbert4:08

      you, TSMC. Any change to what TSMC is doing now if you became CEO tomorrow?

   2. BTBen Thompson4:15

      Not really. It's a very complicated [chuckles] situation, to say the least, for lots of reasons, not just political, but technological. I think probably doing what they're doing. Broadly, it makes sense. Yeah, I don't know. I've written a ton about them and where they are, and you could talk about maybe some of the pricing stuff. They were pretty slow to raise prices, even in the face of shortages. They've been much more aggressive about that over the last little bit, which seems reasonable to me. It is opening the door for when and if a competitor comes along to undercut them in that regard. I think they're probably blessed by the weakness of their competition, but the long-term risk is obviously, number one, just the geopolitical risk. The reality is, is TSMC, a core to their model, essential to their model of being so flexible, of being so adaptive, of incorporating new equipment, of accommodating hundreds and hundreds or thousands of customers, is all their engineering's in one place, and that place is Taiwan, and that has geopolitical risk. And from TSMC's perspective, I think that's just a reality. You can't really hedge against that, and so if anything, yes, they're building these fabs in the US, I think largely for political reasons. And sure, do that if, you know, the Taiwanese government feels that's what needs to be done to keep the US on board, keep the US happy. Build a couple in Japan, because Japan is a future ally. But I think you j- you do have to sort of double down on Taiwan and roll the dice that nothing happens. I, I just don't... I, I don't think it's a hedgeable risk, the, the, the China risk. The other risk is the, you know, Moore's law sort of running out. I mean, EUV has another five, six years, and then it's not super clear, uh, what's after that.

   3. BGBen Gilbert5:54

      How can you get smaller than one, Ben? You just have one nanometer in this. [laughing]

   4. BTBen Thompson5:57

      Yeah, there is visibility to under one, but-

   5. BGBen Gilbert6:00

      Isn't it, like, two molecules wide or two, what is it, two atoms?

   6. BTBen Thompson6:04

      Uh, it's very small. I'm not sure if it's quite that small-

   7. BGBen Gilbert6:06

      [laughing]

   8. BTBen Thompson6:06

      ... but it's ridiculously small. So obviously, I, I would imagine they're investing heavily in-... they've been doing this, like advanced packaging, like y- y- multiple chips on, you know, chip- the chiplet sort of approach. AMD is doing that. Intel's now doing that, and getting really good at that stuff is gonna be super important. I mean, I actually, one thing I do like about the Japan investment is they're trailing edge. I think they're 28-nanometer fabs, uh, which is really where China is making a lot of progress because there's no real economic reason to build trailing-edge fabs. The whole idea is you build it once, and you, and it's long since depreciated, and you're still making money selling cheap chips, uh, out of, out of that fab. But I don't know. Maybe there's something in the trailing edge.

   9. BGBen Gilbert6:45

      More things need chips, and they don't all need the leading edge.

   10. BTBen Thompson6:48

       That's right. The economics of it are very difficult. I mean, the reality is that's why China's filling the gap, 'cause that's all they can really make, uh, economically, and they want... And you, you, you don't just jump to the leading edge. You have to sort of build your way up, and so China has a motivation and an economic rationale, and a lot of the chips that they're manufacturing go right into products that are already made in China, and so they're filling in that gap. You know, that's something that would be beneficial, but at the end of the day, TSMC is Taiwan, and that's gonna always be the biggest risk, and I'm not sure there's really much they can do about that.

4. 7:18 – 9:20

   Amazon analysis

   1. BGBen Gilbert7:18

      Okay, last one, Amazon. David and I did seven hours of amazon.com and then a big AWS episode, and our, I mean, at least my big takeaway was, it is Day 2, and Day 2 is about, uh, becoming a profitable company where the big sell story of tomorrow is realized today, and you gotta lean into that, and that means lots of changes in how they organize, how they innovate, and what markets they choose to enter. And I'm sort of curious for your take on, is Amazon a Day 2 company? And, um, then the same question: if you were to become CEO, what would you do?

   2. BTBen Thompson7:51

      I think there's an analogy to, like, real life. When you're young, you're like, "I'm never gonna become an old fogey like, like those old people." And then you get old like me, and the people around you that are trying to still be young are just kind of pathetic.

   3. BGBen Gilbert8:04

      [chuckles]

   4. BTBen Thompson8:04

      And actually, being old is kind of great. My kids are fairly independent. They can take care of themselves. I can go hang out with friends. You know, obviously, I have more means than I did when I was younger, and I'm a big advocate of just in general living in the present and embracing who you are and where you are in your life stage, and I think that's good personally, and I think you're spot on. That's also true from a company perspective. You can't be a startup forever, and it's bad, too. I wrote about from Day 2 to one day or something about Amazon when Bezos sort of, like, reasserted control a few years ago. He was like, "No, we're gonna do one-day delivery, and there's been too much time spent trying to squeeze our suppliers for profits and margins, and no, we need to get back to this." And in retrospect, that was the seed of Amazon's kind of disastrous last few years, which was they dramatically overinvested in their logistics network. Their cost structure got completely out of control. They overhired, and it's probably a good example, and maybe, uh, it was a precursor of what we saw with Jeff Bezos personally, of sort of losing track of where you are in life and what actually makes sense and trying to be young forever. So I agree with you broadly, um, with that bit about Amazon.

5. 9:20 – 12:18

   AWS analysis

   1. BTBen Thompson9:20

      Do you think that applies to AWS, too, or is that different? Realistically, the challenge for AWS is, number one, what's gonna happen with the, you know, real dry up in startup formation and, and easy money in that space, a lot of which went to Amazon? Um, they're sort of the default choice for startups. Uh, number two, you know, Microsoft's sort of bread and butter has been, "Look, you've been working us for a long time. We're gonna package it at, like, suddenly you're not just paying for on-premise Windows. It also includes Azure credits, and we're gonna be able to attribute that to our numbers, and now you can sort of move pieces over, and we'll help you do it," which I think is, is exactly what they should do. I, I think it's, it's been a very smart strategy, but I think that the issue in any market is... It's like me. All the initial easy stuff is, like, there's lots of low-hanging fruit, but then the largest part of the market is still the part of the market that's sort of always been there and isn't necessarily fast-moving, and it's sort of, you know... And Microsoft is just really cleaned up in that market. Like, like, like, "We're, we're the known entity. We can help you move over," and I think from Amazon building up the support capabilities and rationalizing their offerings... I mean, it's weird because Amazon benefits because they have so many features, right? Amazon, AWS is the Microsoft Word of cloud providers- [laughs] - in that they have an absolute absurd number of features, and the interface is pretty terrible, but every single customer is completely dependent on one of those features. [chuckles] And if that feature is not there, then they gotta go away, and that actually ends up being their moat.

   2. BGBen Gilbert10:50

      We were researching the episode, and we were talking to longtime AWS veterans to sort of understand the mental framework to talk about them today, and that became very clear that AWS doesn't deprecate features. Amazon will kill stuff, kill a Fire phone, kill local delivery for food. They'll kill all kinds of stuff, but in AWS land, if a customer is depending on something, AWS's long-term enterprise value is determined by customers believing that Amazon will continue to support them forever. They don't deprecate services, even when those services end up flipping upside down on the unit economics, and they have to, like, maintain a costly service that they never figured out how to optimize.

   3. BTBen Thompson11:26

      Right, and every time a customer does something custom for AWS, like, it just, it's lock, lock-in, right? I mean, everyone fantasizes about this world where, you know, like, uh, you know, you... Every time it comes up, whether it be, you know, containers or you have all these, you know, IBM's, like, talking about doing this when they acquire Red Hat. "We're gonna make it so you can be cloud agnostic," right?

   4. BGBen Gilbert11:48

      Right.

   5. BTBen Thompson11:48

      And then it just turns out that, well, cloud agnostic, but this one little piece would be, would be better if I could see that service, right?

   6. BGBen Gilbert11:55

      [laughs]

   7. BTBen Thompson11:55

      And then you wake up, and, you know, of course, I wanted Passport to be cloud agnostic. Well, Passport's not moving off AWS, [chuckles] like, anytime soon.

   8. BGBen Gilbert12:02

      [laughs]

   9. BTBen Thompson12:02

      I can promise you that. And so, yeah, it's the Microsoft strategy. Mm, A- Amazon has a lot of old Microsoft people in it. A ton. It's really interesting to consider, like, the different cultures between Seattle and San Francisco. Seattle just is a platform town. Like, that, like, Microsoft

6. 12:18 – 14:09

   Microsoft and AWS

   1. BTBen Thompson12:18

      was the, was the originator. Microsoft's always been the pla- best platform administrator. All that backwards compatibility that you wanna make fun of from a user perspective is essential to building this foundation that people trust, and that locks them in, and they're happy to be locked in because they don't wanna go change it anyway, right? And Amazon does this with AWS. Obviously, Microsoft is doing that with Azure, and the Silicon Valley companies are just very, very bad at that, [chuckles] right? Like, no one trusts Google, right? [chuckles] No one trusts Facebook. Silicon Valley is much more consumer-focused. Even the SaaS companies, those are consumerized enterprise technology. The whole idea is, like, "No, w- don't worry. You never need to pay, pay for an upgrade. We're upgrading on the back end all ourself," but that means, like, they will remove stuff because it's, it's one thing to remove an API that a piece of software depends on versus removing a feature that, yeah, maybe your customers are annoyed, but it's not actually breaking, like, what they operate on. And it's just pretty interesting to see those differences, which I do think there's a geographic aspect to it. I mean, geography is what saved Microsoft because when Microsoft was in the dumps, if they were in Silicon Valley, all their best talent would have left and gone to work for other companies. But all their best talent had kids. They had, they had families. They didn't wanna work for Amazon. Those people are maniacs, and so they stayed at Microsoft. [laughs] And they were miserable, and they bitched, and they wrote snarky blog posts about the company, but then when Nadella came in and sort of refocused the company, they had this foundation of talent that, uh, the HPs of the world, the Yahoos of the world, had, had long since lost.

   2. SPSpeaker13:50

      [singing] Who got the truth? Hmm. Is it you? Is it you? Is it you? Sit me down, say it straight, another story on the way. Who got the truth? [upbeat music] Who got the truth now? Hmm.

Episode duration: 14:09