Asymetric

Horace not having long arms, shares his thoughts the un/bundling of LLMs, and opines on Korean stock market turmoil.

Available to Asymco One subscribers at asymco.com/podcasts/.

Audio Recording

https://asymco.com/wp-content/uploads/2026/08/asy-139.mp3

---

Transcript

Judd
Several microphones, several local hard drives, several cloud drives, and yet sometimes. Nothing gets recorded. This is asymmetric.

Apple announced Apple upgrade. Some people are very upset. Other people are loving it.

You've talked about how this was This idea of basically a hardware subscription had kind of been inevitable. Yeah. But it's taken us a little while to get here.

So You you are I've never seen you happier. Tell us tell us why you're overjoyed and why we should all be overjoyed.

Horace
think about Apple products as a some something you pay on a month on a daily basis. Not pay as I should say you should consider what it costs you on a daily basis. And that's a un that's called unit economics.

It's like, what does it cost for a cup of coffee? But what does it cost to own a car? What does it cost to rent a house or whatnot.

So years and years ago I have to give a little credit to to who gave me this inspiration. It was actually Jean-Louis Gasset. And like could like twenty JLG twenty 12 or earlier even.

We had a lunch in Palo Alto, I remember. And you know he's he's Who paid? I think I think he did, yeah.

Yeah, he's a good guy. he said, you know, you should look at Macs and phones and everything as a what does it cost per day? The problem back then and for until last week, was that, you know, you paid a thousand bucks for a computer or a phone or whatever it was.

And the way to calculate what it's costing you a day is knowing how long you're gonna use it. then subtracting how much you're gonna get for it as a secondary, you know, or as a as a used product. So you're gonna trade it in or you're going to sell it to your friendly you know f Facebook marketplace or you're gonna find the you know a a mall guy or something.

I don't know how people even do that. Or you give it to your children or you give it to your family. And so they use it for another year or two, and then you calculate, okay, I've this has been in use for five years now, so I'd have to take my initial purchase price divided by the number of months in five years and th and then and then do all the math.

Well, if you did that on average My point was, based on this suggestion, that Apple products were about a dollar a day. Now that assumed that, for example, you would keep an iPhone that cost you let's say eight hundred dollars, you'd keep it for three years and then get back. m whatever it was like two hundred dollars at the end, right?

So that would be like roughly six hundred dollars divided over two years. You know, give or take, you it comes out to about a dollar a day. But now it's a lot easier because now when you go through Apple upgrade, you'll get that price on a monthly basis.

All you have to do is divide by thirty or m you know the equivalent average month. which is 365 divided by 12. So that is b that becomes your daily cost to own the product.

And at the end, yeah, you could extend its life or your ownership of it. through buying out the lease and the value is whatever it is left over from the original purchase price having you know paid into it with with lease So there's no there's no interest, by the way. That's the amazing thing.

It's like you you get a an interest-free loan, at the end you can choose to pay the remainder. That you owe, or you can just turn it in and get another one and continue. And all of that works out so the the w most expensive option is to have the one-year lease.

and you're paying a little bit higher, I think, for depending again, so depending on on the configuration. Now this is where it gets gets a little bit tricky. So you have to go And what I did is actually to calculate on average, you have to go out there and and pull every configuration.

Get those those revenue or those cost numbers and then decide okay what is the average from all of these price points because some are more popular than others. Now the good thing is that these tend to d what I did is I picked the median, which is the midpoint of all these prices and said, okay, let's call that the most common and call that the average. And if you do that with the iPhone, it comes up to about a dollar 13 to $1.

15 per per per day. So the old rule of thumb of the dollar per day holds even though this was initially determined In 2012, and it hasn't changed since. So I was very surprised.

By the way, this is on the longer lease. Shorter lease is a little bit more money, like a dollar thirty or something like that So anyway, it's but but most people think of it as like, oh my god, it's a thousand dollars or it's twelve hundred dollars or it's uh whatever the Mac might be, you know. And Macs, by the way, are surprisingly also in fairly cheap, like a buck fifty a day because you hold on to them a little bit longer.

By the way the lease terms on the Mac are two years or three years. On the phone it's one year or two years. And the iPad is also two to three years.

Judd
So a couple of things. So right now it's only available in the US. And they've had programs like this before, at least for the phone.

I don't think they've done it before for um for the Mac itself. But for Apple, what's the real advantage? Like what's the they they have the Apple card, so if you you know, you wanted to get basically interest-free loan, you get an Apple card and then Buy it with an Apple card.

Horace
And you get your three percent cash back, all that. You So what's the advantage for Apple for Well a couple of things economically it might be a little bit advantageous because you get the product back with a reliability of like most people are gonna turn them in and and they get to probably resell it at a little bit higher than would have been the residual value. because people have been yeah, so they you know if it's residual 200 then maybe it's worth a little bit more, it's maybe 400, depend again, a lot of variables.

We're talking hundreds of different configuration here so that's one thing is is that the majority of people are probably leaving a little bit of money on the table, whereas they if they paid for it off they could maybe even get a better price on their own if they went and shopped around or or or marketed their product. That's one thing. The other thing though that's more interesting to me as an analyst is that you get a lot more predictability in terms of having access to large inventory of used products, which means for Apple, now the some might be in poor shape and they're gonna get refurbished or disposed of, but there might be quite a few that allow them to market more effectively into emerging markets.

So the problem has been that they've always offloaded that job, you know, like dealing with second hand and third gen. I mean there's refer, but but this is gonna drive that volume a lot higher and then it might allow them to kind of maybe create a secondary market that's a little bit more upscaled than what they have. what people have to rely on today in these markets.

So there's there's something interesting in that secondary market control. And and there's also the data overall. Like you have you know having people on lease paying you every month.

You have that same relationship with with the mobile operators who kind of know the financial activities if you if you will a little bit. And and so inf it's funny actually how they might steal, quote unquote, steal a little bit of that knowledge away from operators, because operators may end up not having as f as many sales through that channel 'Cause you'll get through Apple and it's monthly. But that's why they bundled the operator with the phone.

You'll notice if you try to buy a phone, you have to buy it with an operator contract now. even though it's unlocked. So the assumption is that in the US you're going to still need to go through a major operator to get the phone.

So there's there's something of a of a glitch there but but there's something knowledge this on so number one it's it's gonna be secondary markets are more more reliable for apples and s and and and in addition there's There's probably more information they're gonna be able to mine.

Judd
So just to clarify, so when you say secondary market, do you mean Apple will sell in a used you know pre a Apple will sell a lease return phone in Bangladesh, as opposed to a local retailer or an Apple store doing it. And does that mean that they're not going to expand the program outside of the city?

Horace
No, I think they will. And I think what's going to happen is they'll have They'll have Apple Store Referb for lack of better term, because I think it's used now, but Apple Store Refurb inventory will be much higher. And maybe they'll market them a bit better And then they'll have the the Refurb refurb Apple certified products in those secondary markets or used markets. Um under Apple branding, there'll be a lot more inventory there and it'll be much easier for Apple to then get that customer in the front door in these markets.

Judd
Well, I'm glad you're excited.

Horace
I Well I would I'm so excited I w I would do it this way. If I had probably the next phone I'll buy. I'm gonna do it this way.

And I'm a little bit concerned about that operator deal but but but maybe I'll do it for other products. The watch is available, the iPad, the I the Mac And and I would recommend this over it's it's essentially an interest-free loan and you you still get three percent if you pay with their your Apple card. 3% back, cash back.

Judd
Yeah. This was also the last last week Apple reported earnings. I hear they did quite well.

So congratulations to all involved. This is the last earnings call that Tim Cook was on as CEO. He will very quickly become executive chairman and there will be a new CEO, mister Suns out, guns out himself, John Turnus.

what is the one thing he absolutely cannot do. And I don't mean like a major strategic decision. I mean a very small tactical decision.

Like what is one tiny little thing that he absolutely cannot do as CEO of this company we refer to as Apple.

Horace
first of all let me just say I the the the word is that the way he's referred to internally is JT. So Justin Timberlake?

Judd
No, just John Turn is his JT and Justin Timberlake is JT.

Horace
Oh no.

Judd
I had If Justin Timberlake became CEO of Apple I would be I think that would be just very impressive.

Horace
Maybe maybe I'm wrong about this, but I heard him referred to as JT. So but because I don't know what to say. You know, there's Tim, Tim Apple Cook, I don't know if we're gonna call him John, you know.

John is so generic. Um Yeah. Turnus.

Turnus is also a little too formal. Anyway, um, yeah, maybe it'll be John Apple. Could be.

So what was your question? What is it something that he cannot do

Judd
Yeah, absolutely like if he does this one little thing, I don't mean like a big thing, I don't mean like he decides to put out the Apple car I mean a small thing. What's a small thing he cannot do or Apple will truly be doomed?

Horace
Well, I probably I'm gonna it's a little a little weird question, but I think He cannot go to the Apple Mac Cafe and expect not to pay for his lunch If he walks out the door without paying, I think his reputation will be demolished. I think he'll be ostracized Famously, you know the story of how Steve Jobs did this. He he he somehow always paid you know, if you were if you were a guest.

And and the thing was though that the way it works is you swipe your your ID and then it will deduct from your paycheck the uh the amount and this was very convenient. And the thing was though that at that time Steve Jobs had a paycheck of one dollar. He he was insisting to pay for other people And in so doing I was stealing from the company lunch once in a while.

So nobody nobody bothered to catch him on that. But I think if he Steve could get away with that, but I think John He's not gonna be able to get away with that. He's he's he's gonna have to pay for lunch.

Judd
Has he offered to work for speaking of a dollar, was it a dollar a day or was it a dollar a day?

Horace
A dollar a year actually, I think For a while at least, he pr he you know he came on with that deal.

Judd
I've never seen him eat. Steve or John? John. Well.

Horace
You know, he's a fit guy. He probably eats a lot. He seems to be a swimmer.

I think he was a swimmer, right? He's got the physique. Gotta have those long arms.

Yeah. Long legs. I used to be a swimmer, but I'm not a very good one.

I swam in high school and

Judd
I wasn't very good because I don't have the the long arms. It took a while for Tim Cook to have his his own product that was not just considered the latest version of something that had had been introduced under Steve Jobs. I think it was the watch. Yep Is there we're running out of of things to introduce. does JT need to have his own No I think he will.

Horace
In fact, and I know what it is. it's gonna be the glass the Spectacles, the Apple Vision, not Vision Pro. I don't know what they're gonna call it, but it'll be it'll be the equivalent of of um whatever Meta's name is, but yeah, it will be eyewear.

And I think the eyewear market is very exciting and I think There's been rumors and rumors about what they'll do next. the vi the Vision Pro was sort of looked like a bit early. and maybe the next first the next version will actually not even have any display in the in the lenses but it'll be just it'll just be passive lenses But eventually the lenses will be displays and that category of product, um, although under development But you don't consider that the same category as the Vision Provision Pro I mean Vision Pro if anything was a prototype and yes it is it is effectively the beginning of that category, but it's like saying You know, because the iPad was developed before the iPhone, and and yet the iPhone is reco regarded as the sort of the defining product for multi touch and all of that thing, all that followed with the iOS.

and you don't you interface. But spatial computing is what we're talking about at a software level and as an interface level. And spatial computing is gonna carry on and be embedded in other things.

But but I think the the whatever they're gonna be called, the Apple Vision it will be the product that most people remember. Remember also the Model T was not the first car that existed and certainly not even the m first car that Henry Ford built. I mean the you know so we tend to remember version three of something usually as the the sort of the the actual s genesis even though it is it is it is really built on generations of other work.

Judd
But it was the first one that people could afford. That could be mass produced.

Horace
Well, if even the models that Henry Ford always developed were a little bit the probably they didn't work so well, they weren't marketed, they weren't There wasn't a production system for them. And so he he had something called a quadricycle as one of his earliest cars. And that was like I forget what year it was, 1905 or something.

Yeah, anyway, I digress. But but the the the the the work on Vision Pro started seven, eight years before it launched. So we're probably you know already a decade into it now.

Judd
Okay. So you would consider those to be this d distinct products though, the vision and the vision product.

Horace
Well, I think John will be remembered as the guy who introduced the vision. That's how old it'll be. It won't be that Tim oh wait a minute.

Wasn't there a Vision Pro before that and Tim introduced that? No, I think John will be I'm gonna bring it up. Well John I think John will be remembered for this.

This is the thing.

Judd
I wish him luck. Um You posted a YouTube video to your X account. Maybe you maybe you post it elsewhere.

as well. But it was it was a a gentleman, I forget his name, some sort of he is a doctor, I believe in Aussie, and he works at a place called Kira And they are making software in order for people to make their own local AI, most likely using upcoming Mac Studio that'll have all the RAM in the world. And your your only comment was a data center never Oh oh I see.

Horace
Yeah I okay I remember now it finally clicks. That the I thought you meant a video that I made, but it was a video someone else made, yes And that was Yeah, so this guy who I don't really follow, but I have seen a couple of his videos and he's he's a I think he's British or Australian, like you said. And he he he he writes or speaks a lot about AI and you know he s he got excited about the fact that There's a generation of Macs that will come soon that will have unified memory in the dozens of gigabytes.

Okay? Now That means, as he put it, as a data center on your desk, you have multiple processors, you have multiple cores, and then you have multi-this giant addressable memory and so yeah the you know I don't know what that really means but in practice the and also by the way Apple in invested in a company that shrinks the size of models by like a factor of ten or more. So there's there's so the models are shrinking in size and there might it might be soon possible to fit everything.

in a in a desktop. That's why it's interesting. And I think I you know, the the lI sound a little bit old fashioned here, but I do remember when we went through this before with mainframes and other big computers becoming small.

And then the Apple would just rode that wave of my miniaturization, the so-called microcomputer Then it rode the wave of getting even smaller to the pocket sized computer. And now many people are saying, yeah, but these models are so big they can only be be delivered by the cloud cloud data centers that are the size of I don't know buildings big buildings and yeah I just keep thinking the old days you know we we The tendency is to get things to be smaller and s get things to be personal, get things to be cheaper, and uh That's that's where we're going. I do believe, I've said this before, but I do believe Apple is even going to make their own servers.

And their own server processors as well. So even though that future of a data center on your desk will be possible, maybe they'll also have Apple Silicon data centers as well because they you know they'll they'll they'll be doing some heavy lifting in the cloud for Apple's own LLMs. and their own services.

So the the the chances are they they are developing their own alternatives to nvidia's processors So Apple's doing its own data centers for a long time, not always with its own Macs in the data centers, but We'll see more and more of Apple's data center strategy coming forward.

Judd
Maybe clarify that a little bit for me. What if there you know if more and more of this stuff is gonna be if it's gonna be easier for more of this stuff to be local Yeah, what does the data sit?

Horace
Well there's a local model and there's a global model. So Apple's already kind of hinted at this in dubbed up DC. So what they what they've done there is They're using Google on the world models that are in the cloud, and then they're using their own, there's like four, I think, four different um models that are available to to developers and users.

And the stuff in the cloud is is also used to train some of the local models But the global stuff is, let's say, knowledge, general knowledge, world knowledge, world understanding, okay? And it's also up to date, constantly being refreshed. And the problem with the local is that that's where the private knowledge is.

So to the extent that you need, hey, can you index my emails? Or rather index this like the kind of questions you're going to ask your computer to do would be organize my calendar. you know connect to my friends and contacts, do things with my personal data that that I need done and so that gets processed locally.

But but but then you ask When's the next flight or how do I connect to databases? Those are gonna be in the cloud. And so the the the the the Google models are better at that world stuff and then Apple's models are better at the private stuff.

and private cloud compute is what Apple's bridging these with. So to the extent that they send information to Google to say, can you fit this data into the global data that will be protected against Google's indexing of it. And so this is why a lot of people were like suspicious that there handing over the keys to Google.

There's a lot of suspicion about Google being the brains behind everything Apple does on AI. That's not the case at all. There's a lot of heavy lifting on on Apple's side as well.

In fact, some suppose you know suppose you are to launch in China, for example, as as Apple will and has been actually authorized to do. They're not going to be using Gemini in China. Or the or in India they're probably not going to use Gemini in in many parts of the world.

So the question is, th there's a plug-and-play architecture where they can they can substitute the world models that fit the markets, right? And so the they've they've architected this whole thing so that there is there is a lot of modularity, I believe. so th that's why I think it you know and also the access to these things will be through through possibly through agents and apps, which are third parties.

that you know will will be yet another level of integration.

Judd
Are these large language models in the process of being bundled or being unbundled?

Horace
Well, they're bundling the global knowledge, and then the the commoditization is happening simultaneously where other open source models are unbundling them So that the this is all happening simultaneously. And this is one of the things that's perplexing is that a lot of Chinese open source models are essentially scraping or copying the the the pr the so-called frontier models and they seemingly are good enough because they're just one one step behind and as as the the As the best models move forward, the open source models are like like right behind. And you're dealing with like a one maybe a one-year lag, maybe less.

And w in which case, I mean last year's AIs were pretty damn good. So if you can get that for free, a lot of people are saying, why am I paying, you know, so much for the latest version of Anthropic or or whatnot. So I think that's the commodization angle is a modularization, somewhat.

And then Apple's gonna just say, well we treat all of these as modules because we think of these as as as modules that we plug into our operating system. And and for that reason I think there's there's this is why there's the whole economics of AI is is cracking or there's a lot of confidence the being uh lost by by by the s by the hyperscalers and and if their economics don't work then they stop spending on data centers. Then then the whole semiconductor bubble bursts And and the data center bubble bursts.

By the way, I don't know how much you've been paying attention. There's a lot of chaos going on in in Korea with with that market. The stock market there went through Girations up up over 100% in a couple of months, now crashing like 30-40% in a couple of days and then up again fifteen percent the next day.

It's tripping all of these circuit breakers almost every other day. And there was a the th this was done the this is ha the tr the circuit breakers were being tripped once every ten years and now it's like every ten days.

Judd
Circuit breakers in terms of the market being shut down.

Horace
So the you you know the market opens. The market opens and these were circuit breakers were established precisely to protect let's say I don't know, the small investor or or whomever. But it was if it drops I think it was there's a threshold, I forget what it is, maybe five percent.

So if it drops be beyond that, then suddenly the the market stops. You know the biggest thing, but trading ends for for a while until the O starts again. And and this is causing not just pain, but the problem is that a lot of the people who are investing in the Korean stock market are small investors, retail investors, mom and pop investors And they've been getting in the game while the m market is booming, and then they've been leveraging themselves with so-called single what is it single equity ETFs, exchange traded funds that are trading just one thing so that they can get leverage.

So people are, you know, four times exposed to to something. But you're four times up and four times down. So you're you're They they get margin called.

And so what happened is millions, literally millions of Korean small investors have gotten wiped out in like the last week and and various estimates out there, how many people are were affected and it it's it's it's scary and it's all because semiconductors, as I said which is including Samsung are being bit up to crazy trillion dollar valuations and then the rug is pulled. and and I think it's it it's one of the aspects of the mania that is driving also Apple's component pricing up, but it's actually causing complete societal level disturbance in in the in a major economy right now. So we're we're we're you you know we're talking about potential for trauma and maybe even a a recession, maybe even a a a real estate crisis.

One of the funny things. Yeah, in Korea. So far so far it hasn't spread yet, but you know, it could spread to Japan.

Japan's having its own issues with the currency there. The yen is being is is being bailed out and under pressure and there's all kinds of things. We could be looking at m at at bank runs, we could be looking at a lot of things.

Russia is another case in point. the oil market is in chaos. But the what I want to point out about Korea that's particular is why so many mom and pop investors exist is because As I understand it in Korea, if you had a landlord, you're not allowed to to own you're not allowed to charge rent.

What you do is you collect a deposit. from the renter and then you're s you're supposed to give it back to them on the lease end fully. And so people who rent in quote unquote rent in in uh in Korea really don't have enough money to buy, but they may have like twenty or thirty percent of the value of the house, so they'll hand over hundreds of thousands of dollars equivalent as a deposit and the only way that the landlord makes money is by investing that money to try to maximize their return and that it should be higher than what you get in rent.

Now think about that Because then that that encourages speculation and encourages precisely these sort of investments in crazy in crazy leverages. And and so these landlords who might be doing this and they get margin called, then they lose their house because the bank says, you know, um This is only only collateral or the only capital you have. So I don't know exactly.

I'm just saying that this this is scary. AI is not just An interesting odd technology, the way it's being and one of the big lessons here is if it's a bad business generally speaking, right, and we are not sure it is what it is right now. It's it could be good for some, bad for others.

But the ide the problem with misallocating capital is that it's not as it's not a victimless crime. It's like saying, well, the investor lost. But when y when society does it on such a grand scale, when with capital is so badly allocated, markets are supposed to be really good and efficient and really smart about allocating capital, right?

They make mistakes, but the the damage is contained. and and the the investor learns pro supposedly. But here we have something.

And I wrote this piece, Why is everyone all in on AI? and m my problem with this is that they're all in even though we're not sure it's a great business And if it isn't, then then it it it's maybe the first equity led financial crisis we've ever seen, because everything else has been debt driven. I mean ever, not ever, you know, the initial investments in by you know in peop the famous Dutch pul bulbs that were tulip tulips, yeah That that would that was effectively also like an equity an equity.

Judd
I'm a little rusty on the Dutch tulip crisis. Um in the from the sixteen hundreds. tie those together for me.

Horace
How how make What I'm trying to point out is that there's typically when you get to s what we call typically a um Not just a market crash, but it's contagious enough so that an entire economy, globally possibly, ends up in a recession. So do you have a global crisis? So we have the last was 2008.

And that was driven by bets on on i e derivatives, but ultimately real estate exactly. And a lot of those were mortgage backed securities and and a lot of those were fundamentally tied into the bond market. It wasn't that it was any stock market bubble.

It was it was it was something in a whole different type of equity. I mean that that the stock market got dragged down by a crisis in debt, let's put it this way simply. And if you go back and look at other crises It's been usually behind the real estate like Japan has problems I mentioned Korea, but also China.

China has huge real estate problems. And because the amount of debt out there is far bigger than the amount of equity, the debt market is far, far, far bigger. Than the equity market.

So the equity market, if you were to sit all up, is you imagine all the businesses that are traded on all the stock exchanges in the world. Take all that market capitalization, add it all up, it wouldn't be one percent of the debt market. Because everything in real estate is debt.

Everything in commercial real estate, private real estate, then you have commercial loans to to you know businesses that are f you know underwriting things like factories and other things though that may be capital in capital expenses. But it there's real estate is by far the biggest asset class, even in amongst let's say all the wealth in the world. More than I'd I you know, I'd guess equities make up probably fifteen percent of the wealth in the world.

And and and so everything else is debt or or cash. And and and so then or other things. I don't know, the art market is tiny, but but the y you know, I can't overstate the importance of a stable debt market.

And this is why when even when Trump was messing around with tariffs, he had to pull back from some statement or p policy because he quote quote unquote, you know the the the bond market was getting jittery. I used the different phrase, but it was essentially the bond market was and if bonds by the way, the bond market crashes then you know, treasuries are in trouble. The m the US can't borrow anymore you know, there's there's issue issues with with the federal federal government liquidity.

So so there's there's it's profound and what what I'm saying is that when you talk about contagion There's a famous phrase or s word. It says that the the the bad debt is contagious. Well, what means is a the bank fails And because other banks have pieces of that bank or the assets in that bank, they're all going down start to you know domino effect.

So the the thing about AI so far is that it seems like it's not apart from the hyperscalers. It's not likely that it's going to be contagious to other parts of the economy if things go bad. But it looks more and more, and it's always surprising because you don't know where it's coming from, where the threat is Sometimes these things are hidden from view and they're deliberately hidden because they're nasty.

So here the problem we have with AI is is a circular is circularity in terms of relationship that Nvidia is paid by the hyperscalers but then it turns around and finances the hyperscalers. Right. Finances and owns equity in OpenAI and Anthropic in in this or that whatever project is involved, startups and so on.

And and so you get this process of knocking over one another. So you know I'm just warning. I'm not an expert any of this, but I'm just warning that The bigger it is, the more likely that there's stuff that you cannot see as far as the the dirty laundry.

Judd
Well I'm worried about I'm worried about these Korean landlords. What are they gonna do?

Horace
Yeah, and I I've I have one friend from who from from Korea and I'm sometimes pinging him and asking him are you alright? Are you on a ledge somewhere? Um But I try to I try to to get some inside info because I'm I'm just reading reading the headlines is not the You know, there's a there's there's a lot going on on the street level.

So Korea is a weird place, you know. You also have North Korea, but you also have you have politics. Yeah, but the the the whole c I've been there a couple of times.

The whole it's very different than uh Then what you might expect. I mean in many ways.

Judd
Rent thing is I don't think I've ever heard of that before.

Horace
Yeah, but plus there's other things like like the government interferes a lot in the Chables.

Judd
The Chables are the large organizations, like conglomerates.

Horace
Conglomerates. So what happens is like they're not allowed to participate in certain businesses there, you know, because they don't want the small businesses to be crushed. So one example is I was there visiting and I was talking about micromobility and I said hey why don't you let's say Samsung or LG I'm not gonna name who it is but it was one of the Chables And I said, why don't you guys build e-bikes or or scooters?

And they're like, we're not allowed to because bikes are considered a small business. that local people are running and if and Samsung is has a gentleman's agreement with the government that they're not gonna go make bicycles. I mean that's a that's a sort of like informality but also like there's there's this this and there's a bit of socialism if I may say because I think there's a lot of a lot of Koreans are kind of Quite, although it's a very capitalistic country, there's a lot of social justice d demanding.

You know, people are very kind of w the workers' rights and this and that and there's I suppose there's a there's a left and a right in politics and the left is swinging sometimes w you know, one you know, in power and others there's a reaction. And so when when they had this gr constitutional crisis it was exactly because like the right was was one fomenting for for, you know, a coup and stuff like that. So it's a very is is I don't know what to say, but I would say that in it sounds to me like Korea is a little bit A little bit of a canary in the coal mine here.

So maybe maybe you should keep an eye on it as as this AI story develops.

Judd
Yeah, no, for sure. Can you briefly I think I have a decent understanding of what a che ball is, but maybe you could explain it more succinctly than I could.

Horace
So the Che Balls are similar to the Keiretsu in Japan. And Kiretsu in Japan were informal or formal relationships between so it's not quite as you said a conglomerate but it's let's say a big company is in a partnership with another big company and they form together Not a cartel because they're in different businesses. So Toyota will be in a relationship with Denso.

Denso makes parts, Toyota makes cars, but somehow Denso is nearly exclusive to Toyota. And then Denso then gets into a relationship with a chemicals company and another relationship with a steel company and so on. And then therefore Toyota is linked to all these other companies and they all kind of are happy together and doing business together but not really a classic cartel.

And so Che Ball is similar, but in this case, maybe Samsung has divisions doing all kinds of things, and that that is a conglomerate. But also is a relationship with, I don't know, they're big vacucleaners, but they're also like, you know, have a relationship with a rental company or something. It's funny Samsung when I looked it up started as a rice trading company or noodles or something.

It was it was food and it was a it was a it was a it was like a wholesale or intermediary. and trading and doing random random things and and over the decades it just became um it just became a a a m a ma mass of business.

Judd
and and I think they were tolerated but isn't there something isn't there like a there's a family component to this. There's something And most of them are family do the there was a original

Horace
Because these are all old post war. They're not that old. We're talking since the fifties probably.

so the the founder or his direct descendants are still alive. And it they tended to be majority but then maybe minority shareholders, but still a huge chunk of the of the equities held by by a family and it gets ugly because again you got this social y you know, resentment against the ultra rich and and and and then they have to they're sometimes accused of all kinds of uh economic white collar crimes and the msometimes they go to jail. The the the the their behavior is scrutinized for signs of insensitivity.

I remember there was a the the daughter of the f president or maybe part owner of the major airline in Korea had an incident where she was flying first class and she she complained to the steward that the service was inadequate, like but she s was so violent and angry about it that You know, he was almost like injured. And it it bubbled out of this, you know, this incident came out because of this. And then she ha Man, she was almost she w they almost lost everything because of that because everybody turned against this.

You know, they'd like canceled her in a sense. But i it's a strange strange thing where you have these well I suppose wasn't there a movie Crazy Crazy Rich Asians or something like that, but that was set in Singapore Mm-hmm. It was about very wealthy people.

Good movie.

Judd
That's a it's a it's a perfect airplane movie. That's a perfect airplane movie. Okay, so basically, uh, in order to keep track of how the the so-called AI business may bring us all down Download and listen to every single BTS alband we'll we'll have a clear understanding.

Horace
That is your expert advice. Correct. Fascinating also Korea becoming a a a an exporter of of soft power and cultural goods, which I always think is an interesting indicator.

Because Japan began that too with uh Comics with cartoons, with with films, famous famous movies in the fifties, Godzilla, all those things from Japan. manga or whatever the comics are. Manga anime.

Yeah anime. But the but I think w again Korea many ways replicates whatever happens in Japan because they they copy what's successful and they decided hey we should do that too so they said hey we can do pop music And then that we can do it with boy bands and we can do it with this. So they you know, K-pop came around and and that to me I would without disrespect, but it just seems a little bit manufactured, you know Like like it's a business business development.

Judd
Yeah, I can't I can't speak to the music part of it, but I will say there are some of the most interesting and exciting filmmakers are Korean.

Horace
Nowadays, yes. Nowadays.

Judd
Yeah. Yep. Yeah. And I ca when I was growing up that really wasn't the case. But I'd say you know fifteen, twenty years ago. For me, there's a movie called Old Boy, and that was the first one that really kind of broke over, and now it's

Horace
well didn't they get an Oscar for what was it Several Parasite in a one one

Judd
several Academy Awards. but it's not uncommon to see in Minnesota at the movie theater cor you know, movies from Korea in Korean

Horace
and Unheard of, unheard of, unbelievable.

Judd
It's it's wild. So kudos, kudos to the Koreans, but man, I hope they figure that out because we want more movies, we want better movies

Horace
So we don't want we don't want Korea to go down in flames and they won't make movies anymore. But I would say it's also by the way, they're living next to an existential threat of North Korea, nuclear nuclear armed f m the second third largest army in the world. And and and and yet they're like they're like super cr super productive, creative, and And apparently super willing to gamble on on on very very very precarious bets on the on the stock market

Judd
Well every day is precarious if you're living next to insane an insane person with nuclear weapons.

Horace
My my questions when I go there is like aren't you guys worried and like well not really but you know if you were in Taiwan and you look at China same thing it's like existential.

Judd
Well hopefully everybody in South Korea and Taiwan listening is feeling good about their prospects. And Horace, anything we're we're coming up on our our time. Anything you'd like to add before we before we bid adieu?

Horace
we're gonna post a b a few videos I believe from my office hours. So if you want to check in on on some of the some of the details of the quarter, but it was a it was a it was a surprisingly good quarter but also there was a surprisingly bad guidance and that had to do with with shortage of components and and so th there's a lot of m madness in the markets right now but uh But overall demand demand for Apple's products is is through the roof. It reminds me a lot of COVID when things were looked very, very crazy and yet people just couldn't get enough Apple product But this time there's there's there's a there's actually not enough supply. So anyway, interesting times. but stay tuned and we'll talk about it all.

Judd
Horace, thanks for the time. Thanks to everybody listening. We will see you again very soon on asymmetric.

Now in Beta for Asymco One Subscribers. You can now listen to Asymetric from your favorite Podcast app. Sign in to Asymco, then visit Podcasts page.

Tap or click on the button corresponding to your choice under Asymetric. This should open up your Podcast app with a private feed for Asymetric.

Should you have any issues, please send email to support@asymco.com.

Horace on WWDC 2026, John Ternus’s timing, what we can learn from the candy bar, and Judd’s indifference to origami. Also, World Cup predictions.

Audio Recording

https://asymco.com/wp-content/uploads/2026/06/asy-138.mp3

---

Now in Beta for Asymco One Subscribers. You can now listen to Asymetric from your favorite Podcast app. Sign in to Asymco, then visit Podcasts page.

Tap or click on the button corresponding to your choice under Asymetric. This should open up your Podcast app with a private feed for Asymetric.

Should you have any issues, please send email to support@asymco.com.

Transcript

Asymetric 138, hosted by Horace Dediu and Judd Rubin.

Judd: The World Cup has only just begun, but somehow Horace has amassed 11 yellow cards. This is asymmetric. Horace, why are you being so rough on the refs and your fellow competitors?

Horace: I don't know what to say because I'm not even watching the World Cup. Never mind playing in it. But so, what can I say? First of all, the World Cup is, for me, something kind of very beloved. I used to watch it quite religiously when I was younger, and my family and my father in particular was always a big fan of what Americans call soccer, but the world calls football. And because I grew up in Europe, and I live in Europe now, it was always a big deal. We would kind of predict those first round games, like what the outcome would be, and have a competition with that. And I even went to see, in 94 when it played, I think it was 94, yeah, when the games were in the US again. And I actually went to see a couple of games. Miami and New England or Boston. But the last few years I just haven't been as excited about it. It feels like it's lost a lot of its charm. It's become very expensive. And I don't know if the right word is corporate. I don't think it is corporate, but it has felt more commercial in a way. Which isn't always bad, but I'm just thinking that football is a people's sport, or it's the beautiful game, the game that everyone plays even if they don't have any money. You just need a ball. And many people just don't, barely even use a real ball.

Judd: It's like they make their own. And feet.

Horace: Yeah, you need feet. But when I was hearing that ticket prices were like thousands of dollars and it was so exclusive, it just didn't feel like it used to. And also, the game has become much more defensive because it turns out you can game the game, so to speak. And the games are kind of a little bit less vibrant. But anyway, I'm not an expert. I just kind of don't have the passion anymore for it.

Judd: Okay. Good to know. When we were talking earlier you had said that it's been a watershed couple of weeks for Apple. Tell me more. Why do you think that is? What has been the major life, world, company changing events or series of events that makes you feel that way?

Horace: Well, we finally had, in WWDC, we finally had the kind of, I don't know if catharsis was the right word, but it's some revelation of what people had been hoping for for a long time. Which is, firstly, well, maybe, I hate to say it, but I think it's kind of like two years ago we had anticipated sort of Apple intelligence being something that was going to change everything, or at least bring AI into the world of Apple, and it didn't seem to do so. And now it seems like it actually has delivered on that promise of two years ago. Now, what it did is actually delivered beyond that, but people haven't yet absorbed just what changes will be coming, I think. Because I think, unlike a two-year-old promise, what we're seeing now is an integration that's far deeper, where the operating system is going to change dramatically the way it works for the user in terms of interpretation of their intentions. So I call it intention computing. And it's going to allow developers to participate with AI inside of their apps, and it'll make apps modular for an intent driven experience for the user. So they're reorchestrating the entire operating system and the entire app ecosystem to accommodate intent computing. That's my read of it. And that intent computing has very much been the promise of AI as a new way of doing computing versus touch computing, which has been the consistent interface for almost 20 years now. And touch was going to propagate into voice computing or spatial computing, and now it's intent computing. Now, actually, voice didn't really go very far. Spatial, the jury's still out, and I do think Apple's pushing hard on spatial, continuing to develop that, but it's far more, let's say spatial is far more, I would say, along the lines of what Mac is, whereas intent is much more along the lines of where phone is. What the dream of those who wanted to replace touch was, those who, for example, if you go back to Alexa, if you go back to, what's the Google version of that? You know, just smart speakers. If you go to Siri, even the original Siri, if you go to Humane, which had the notion of a microphone-only interface, maybe with like a mini projector. And even if you go to what OpenAI suggests that they are building, an AI-first device. The assumption behind all of these is that, first, you don't have touch; second, you don't have apps; and that your entire compute is a conversation with some intelligent agent, or the equivalent of a chatbot. And you ask it what you want to have done and it'll do it. By this I mean intent, because it's kind of like you just ask. It seems a bit utopian, because I think that it's not driven by analyzing what people actually do, but rather hoping to replace what they do now with this new way of doing things. Let's not forget that touch also didn't replace the way we were computing before touch. We had applications on our computers, but they weren't apps as we call them today, and they weren't, you know, involving swiping and gesturing. They involved mostly, you know, clicking and typing. So the idea of intent is just, tell me what you want and I'll get it done. And I think, as I said, it's a little bit utopian. It's, you know, remember Star Trek, you know, asking the computer. And yet when you look at what people actually do, they're not in an inquiry mode. They're not inquiring. They're not requesting. They're chilling. Entertainment, communications, they're doing things with their phones which are far more passive, you know, and far more like television than they are interrogatory or trying to be productive with a machine. That's, I think the reason behind intent's utopian vision is that it's created by computer people who think that computers are utility first, and, well, first, second and third. So I find it a little bit, the visions until now have been more or less, you know, not consumer friendly and not consumer accepted. But what Apple's building, in terms of intent, is a layer within the operating system. You can do those things, but at the same time you maintain all the other things that you're already doing. So you'll have the app economy, you'll have the app ecosystem, but apps will also be stitched together into workflows and experiences that would allow you to get things done. So, for example, instead of turning to three different apps to deal with your travel. You'll have an app for booking your flight, an app for booking your hotel, an app for booking your ride to the airport, maybe something more to do with setting your vacation settings in your calendar, maybe dealing with some shopping that you need to do before you leave. Instead, you might get all those things done with a few voice commands, which then will put together, it has to be iterative, so it'll give you back options and you'll go through them, and it'll connect to these apps that you're normally using. And my thesis is that Apple decided not to completely obsolete the app economy because it is somehow easier to replace all that. It isn't. They decided to keep trust and connections with developers and connections with data sets and connections with the trusted entities that are like banking, shopping, all those places you go. Those relationships will be preserved rather than you giving rights to all those things to a new entity that is this new agent. And so for that reason, I think the future that Apple is envisioning is where AI is layered on top of what we already have. That they call the system orchestrator, which routes your queries to multiple AI models but also to the apps that can work with those models and put the solution together.

Judd: That does sound very nice. It also sounds like a lot of, you know, maybe the methodology is slightly different than what's been talked about before, but as you were saying earlier, this is sort of where they seem to be where they were hoping slash promising to be a couple of years ago. But when I think watershed, I think something new. This feels like they're doing some really interesting things, but they're just sort of catching up to where they thought they would have been a couple years ago.

Horace: Well, Siri's gotten better and Siri has caught up to where it was promising to be. I mean, it's not even the way it should have been two years ago. It's where models are maybe as of six months ago. So things have gotten better from that point of view. So Siri has caught up and then surpassed where it was supposed to be. But my observation about intent is that it isn't only, it isn't just we have a new Siri, or we have new image playground, or we have new text creation tools and all of those things that work better that were introduced two years ago. Rather, we have, at the operating system, the orchestrator. And the ability for developers to work with these new Apple foundation models. By the way, there's five Apple Foundation models. Two are on device, three are in the cloud. One of them is hosted with Google. There is training that was supplied, or rather, you know, there was training data that was supplied by Google, and so Google's data was used to, you know, create these models to some degree. Possibly those will continue, that relationship will continue, and there'll be improvements over time also because of Apple and Google kind of partnering there. But one thing to note is that this isn't Gemini, this isn't any of the Google apps, any of the Google interfaces, and the data never leaves Apple's own servers, or what's called private compute. And so none of that was promised two years ago. The idea of the foundation models being updated. The on device foundation models, which by the way require quite a higher level of device performance that two years ago wasn't available. So they've done a lot to enrich this whole thing. And like I said, connections to developer, connections to Google, connections within the operating system, all of this is new. So let's just summarize it as, it's smarter, but it's also far more integrated.

Judd: That sounds great. You also were not shocked at how the market responded.

Horace: Yes. So the market, the stock fell immediately when, let me just put it as precisely as I can. During the keynote event, there were several segments. There was one on child safety, or, you know, safety in general. There was a previous segment to that that was on changes to the core operating systems, broadly speaking, not very detailed and not operating system by operating system the way we used to do it. And then finally Siri came in about two-thirds in, or Siri AI. And to introduce it came Mike Rockwell, who is the new head of Siri AI. He used to be the head of spatial computing, or vision pro actually, and he delivered vision pro and then moved to this role. As soon as he began speaking, and I mean within seconds, you know, they introduced him, he said hi, I'm here to introduce Siri AI, probably between the second and third syllable of Siri AI, the market fell. I was watching it. I was watching it and it had been up, up, up throughout the anticipation. It had been up for a long time before even the event occurred. But during the event, you can see it keep rising. And then, as soon, I think $317 right before Mike began speaking. And I of course blame Mike Rockwell for this. I mean, his outfit was terrible. His outfit as soon as he comes, not cool. I don't know, it was his haircut at the time, but clearly as soon as he began speaking, it felt like straight down and then it kept going and going and going to the end. Over the next couple of days it fell over twenty-five dollars. I mean, it's, but if the drop started as soon as he said Siri AI, is it just that—

Judd: people are freaked out at the name. If it was a different name, do you think it would have happened? What was the signal? Or did it just hit a number where it was far beyond what you've always described as a point. Many people said, you know, the market's got nothing to do with the event, the market's nothing to do with Siri.

Horace: I saw these feedback, as I had posted this on the social media, and I said no, I was watching, it was very precise. But notice that it wasn't based on what he said because he hadn't even begun explaining. He hadn't even done a demo. It just started falling as soon as, okay, now the presentation begins. It's time to sell, which is indicative of the phenomenon that's known as buy the rumor and sell the news. And the thing is that usually you don't know, but some people, you don't know the timing, so they'll say, well, as soon as the keynote begins, or as soon as the day begins of WWDC, people start to sell. Why not? But here they start to sell only when this Siri announcement was finally made. And it was spectacular. I've never seen a better representation of just sell the rumor, sorry, sell the news, because, you know, Siri, it doesn't matter what it is, they're announcing AI, they're doing all the things that everybody anticipated. It's time to sell, and it just kept going and going and going. And then days later people are trying to explain the quote unquote market's reaction, or Wall Street reaction, or investor reaction, or analyst reaction to WWDC, because it had kept, you know, it had fallen so much.

Judd: Will he be allowed to present again?

Horace: Yeah, exactly. I don't know. But of course, when Steve was on stage back in the day, we had calamitous, catastrophic falls similar to this. One of the worst was when he announced the MacBook Air. I have no idea, but I think the stock fell 20% in the next few days. The very first air, the launch of that brand. It's not explainable because it is entirely driven by some kind of momentum and market, you know, markets are not pricing at that point. What they're doing is, there's an anticipation that other people will sell, so you sell. So it's an avalanche of anticipations. And the same thing before, this is an anticipation of buying, so you buy. So it's all based on what other people might be doing, and what other people might be doing based on what other people might be doing. So it's kind of recursive, and it's a news cycle that leads to this kind of, you know, amplification, on the upside and on the downside.

Judd: So, but again, I thought this was... other than that, how was the play, Mrs. Lincoln?

Horace: that the kind of, these play, these types of games are over for Apple. It used to be much more prevalent, used to be much more common, but it became, let's say, less of a game. But it's interesting how the market itself is so hyperactive right now. We've had these international issues, we've had inflation issues, we've had consolidation or concentration inside of the tech sector. And so it just seems jittery. So my point though is, I kind of felt, because it was back to this game, I found it strangely comforting, because, hey, look, Apple is selling on good news. And the more the news piled on, by the way, throughout the week, you know, WWDC was still running. There were more and more events, more stuff coming out, there was more interviews with the executives and the picture of Siri AI, with, you know, the whole picture of the integration I've been speaking of, the relationship with Google, all those got fleshed out and it sounded better and better and better. Most analysts or whoever got their hands on the software said, wow, it really is pretty good, even being, you know, great, some people would say. And there was a lot of applause from the people watching the performance here of the product. That we're gonna need a lot more powerful hardware from Apple to take advantage of all these things. Obviously that is a positive for the business. So maybe that is not what was expected, but now we have a precision about exactly how many current devices can take advantage of the full scale of this improvement, and it's not many, so a lot more hardware will be needed. So my reading is, as an observer, as an analyst, there was kind of, I would say there wasn't a huge excitement during the event, because it was not their greatest event, especially taking so much time on the safety side of things, which isn't, it's, safety is a hygiene thing. You don't advertise that your system is more hygienic than ever. Just like people can't sell cars because they're safer, that is a not a purchase decision driver. So from the defense, yeah, well it works for one, but not many car makers. So it's flat, it's kind of, for me the whole keynote was a little bit flat. Also, by the way, the Siri demos were slow because they weren't faking it, and therefore that is exactly how their action happens. So from a snappiness and entertainment point of view, it was a bit like, it felt like there were a couple of beats off, you know, in the quality of the video. For me as an observer, again, it was like, okay, this is decent, this isn't great. But then as time after the event passed I was getting more and more and more impressed. Okay. The market went the other way. They hyped it up way in advance. They sold it deeply afterwards, ignoring all the interesting stuff coming out of WWDC. Historically though, that's how WWDC has always been sold off because people would always anticipate more than what was happening. In this case, they delivered actually more than they expected. Most people expected, and yet it's still sold off. And now it's bottomed out and it's probably gonna start recovering. But you know, there's a lot of cynicism around, as usual. And that's, again, encouraging to me in a way, because it means Apple works as an underdog, you know, it works better when it's underestimated. And it's great to see the market throwing away good, you know, the baby with the bathwater.

Judd: What else jumped out at you? I mean, not only just the keynote, but there were other events throughout the week and all sorts of other announcements. Was there anything else in particular you wanna touch on, for Tim Cook's last WWDC as CEO?

Horace: Well, a couple of things have been in the background. The news coming out of the analysts following this, the market for phones and computers and PCs has been, and this isn't getting a lot of highlights, but the market, because of the memory crisis, the market for computers is going to be down this year and the market for phones is going to be probably the worst year anyone can remember. I mean, we're going to see 20% drop for a lot of the vendors out there. And some may actually be almost in a state of, you know, economic or financial distress. Like, who, it's that bad. These are the smaller so-called budget phones, the lower price points of Android, okay. They're just not able to get, you know, any memory, and they're not gonna be able to ship their market shares.

Judd: Would now be a good time for us to introduce an Asymco phone.

Horace: Yeah, sure. It'll just be an iPhone with an Asymco logo on the back and we'll just charge double. But the idea of the market being in serious trouble is contrasted with the fact that Apple is, the same analysts are projecting Apple to be growing. Apple grew in the first quarter, projected equal growth in the second quarter, which is ending June. That growth would roughly be in line with the first quarter, which was like in the twenty percent plus. Now this is at a time when most of the vendors are seeing declines, and then the analysts are suggesting that declines will accelerate into the end of the year. So this sounds to me like a catastrophe. So while the market for smartphones is collapsing, Apple is somehow still increasing its volume. Now we don't know if it's going to actually continue their rapid growth into the second half of the year. But because it has had such a great success with iPhone 17, and because there's new product coming out, probably of the, you know, high-end, maybe a new version of air, a new version of fold, a new version of the pro products. And then the anticipation that 18, base 18 is not coming out until early next year means that 17, base 17 is gonna run through the end of this year. All of this is probably going to increase Apple's revenues in the iPhone segment. The NEO is doing amazingly well.

Judd: But time out, timeout. So did you just say you don't think that the iPhone 18 is gonna come out in the fall?

Horace: Well, that's what the rumors are saying, and they've been saying this for a while. The iPhone lineup into two. Now they already have an iPhone, the S, sort of the low-end version of the iPhone, was a spring product, springtime. So that started with the iPhone, I think it was C back in the day, and then became the S. And now the idea, I think, is that they're actually going to take the standard base model, okay, which is now maybe the mainstream model, and move that to a spring launch and keep the, by the way, there's a reason why in the December quarter the mix of the high-end product is far overwhelming. There's a lot more of the premium stuff sold, and not many people are rushing out to get the latest of the mid or low range of the iPhone. And they know this, and they've known this for a while. So it also makes sense to move around China a little bit. China holiday is in first quarter. And so moving the 18, and possibly also for India, I don't know when the Indian holidays are. But if we can move something off into the spring, it might be an interesting split. Now I don't think it's gonna affect sales that much because, like I said, there's not only a bias towards the high end, but we're going to have more high end than ever before. If a fold is coming, then the high-end lineup will be the pro, the pro max, the air, and then probably ultra will be the branding for the, so we're looking at four high-end iPhones we'll be launching in the fall, with hints of what might be the 18 base model coming in later. And thus also maybe concentrating a lot of the buyers to saying, okay, since this is what the latest is, we're gonna go with one of those four options, driving the ASP higher, and so on and so on. So yes, it's a radical departure, but it's been rumored for at least, I'd say, a year, maybe two years.

Judd: Well, I guess a couple things. I mean, in American football, not soccer, but in American football, there's a saying that when it comes to your quarterback position, if you have two starting quarterbacks, you really have no starting quarterbacks. So if you have four premium phones, don't you have zero? I mean, like, they can't all be premium. They can't all be. They're just expensive.

Horace: Well, you have a good point. And historically, you know, Apple's been about a simple and clean and clear product portfolio. You know, famously when Steve came back he eliminated half a dozen, whatever, there were so many different versions of back. He eliminated a lot of them and then introduced the iMac and it became, as you say, the quarterback, the leading quarterback. But with the iPhone over the years it's kind of gotten a little bit broader. And you see this in what, the so-called, my stairway to heaven is my, the breadth of the products available. Of course that includes different memory configurations, so that there are far more now than there used to be. Used to be one iPhone or maybe two, and now we have like about 15 with all the memory configurations, and not to mention the color configurations. But as far as the so-called premium, now there's this puzzle around fold, and I must say I'm a little bit confused by this product. I think we talked about this before, but I think when it opens, it kind of starts to look and feel a lot like an iPad. Like an iPad mini actually. And I'm not sure that people want to carry an iPad in their pocket. Of course, when closed it's a phone and it feels fine, but it is, the iPad is not necessarily, maybe it'll become, because if it is with you all the time, it'll become far more attractive for you to interact with the iPad type of interface. But I will say that one of the rumors about why they're doing this is that there's a lot of interest in China. And it isn't for software reasons, but rather because it is such a, it's such an event, it's such a theater, it's so theatrical, right? When you have a machine, a moving object, it's a different thing than something static, and so there might be a preference for something that shows action, that shows clever mechanicals. And that could be one of the reasons that market is appreciating it. I'm reminded, by the way, of the debate back when I was at Nokia, that there were flip phones that were very popular in the US. And Nokia resisted going to a flip phone. They had something called a monoblock, or the candy bar. The candy bar phone was the way Americans called it, but inside Nokia we called it the monoblock, and it was better by every possible thing you could measure. It had more durability, it had longer battery life, it was more durable, it was easier to use. You didn't have to do something to get something done. And it just was cheaper too. Obviously you could make it fancy and expensive, but it was cheaper to make and cheaper to manufacture and so on. So everyone, all the engineers were scratching their heads. Why Americans preferred flip phones. And Motorola launched the Razr, which was a super thin flip phone, and that went on to become the number one world bestseller in the phone space. So much so that, you know, they make dozens of versions, and eventually they relied so much on it that when it began to decline the whole Motorola failed. I mean, the whole company essentially collapsed. Strange story, but you know, you have a hit product and a couple of years later you're out of business. That was the story of Motorola. But it wasn't just because it was a fold, it was because they, I don't know, they made a lot of mistakes as well. But flips were very popular. And so when we try to find out why people wanted a flip phone, it was that this mechanical action felt like, I don't know if people felt like, well, that's one idea, but people felt like it's much more satisfying. It's satisfying to open it and answer and close it, to hang up, and especially if you don't like whatever the conversation was, it's satisfying to just hang up on something with violence. So it was, you know, as some people even in Japan, it was common to have tiny little flip phones with little antennas. So you had to flip for the phone and then extend the little antenna to make it work. And that's outrageous, you know. You have simplicity on one side, you have complexity on the other, and people are voting for complexity. Now, I don't suggest, and this is one of the reasons I think people reject, and Nokia, or sorry, Apple rejected a lot of these moving origami, you know, have your laptop flip inside out, have it be an iPad or a tablet with a pen. Do all these things as if it's kind of a convertible. But here we are. So maybe after, you'll know the answer to this question.

Judd: I won't. How popular are the current folding phones from, you know, Samsung or the other manufacturers? I mean, are they that popular?

Horace: Not that popular at all. They've been doing it for years.

Judd: Yeah.

Horace: Partly the reason they're not popular is that they are flimsy, and there's a visible crease, and they're gimmicky and cheap. The, but cheap feeling, but very expensive to buy, and this is the problem. And it isn't a killer function, what do you do when it's open? Well, you have a bigger phone screen, so maybe you watch videos or something. So it hasn't come with a killer app, if you will, for the bigger screen. So this is what's interesting about Apple, is whether they will, first of all the form factor is slightly different. It's shorter in height, and then when it opens it's much more like, like I said, like a mini iPad. And I think once it's in iPad mode it's going to be probably usable for entertainment and other things in a much more intuitive way. Maybe the keyboard and other things will be easier to use. I don't know what they're going to introduce in terms of software, besides just making it an iPad. So from Apple's point of view, they are also getting rid of the crease. It probably mechanically will be nearly magical. Probably they'll solve a lot of the durability issues as well. So there's a lot that they can fix about those products. And then there's something that needs to be done on the software side to make it more attractive and usable. So yes, there's room to improve, but it's not clear that people are craving this, except again, maybe culturally for various reasons. I do believe that $2,000 will make it very, very niche. You know, it'll cost more than a MacBook Pro.

Judd: This just feels like, I don't get it. I don't get it. This feels like it's like the cube. You know, like, sure, you can have a beautiful acrylic cube. But, okay.

Horace: But this is why I don't like to speak about rumored products, although there's a lot of evidence for this one. But no, I brought it up because I think the portfolios, the split between the two. I mean, we're talking about watershed, okay, and we're talking about how the market is contracting for phones. Into this, Apple is expanding its portfolio. It's going to launch a lot of new product in the fall. There's a lot of expectation that they might, if not volume wise, but at least in dollar wise, going to expand, they're gonna grow their business. And their market share probably will increase substantially. This is where it comes down to. Historically, Apple's had 18 to 20 percent of the phone market. A lot more in terms of the install base because the phones last longer. But with this crisis in memory, and with this market expansion in product, the lineup and the new spread of launch dates, it's possible that Apple is going to reach 22 to 25% market share in devices, or, I should say, in iPhone. Okay. In the phone market. And at the same time, in the PC market, we're looking at 10 million NEOs this year. I think the market share in the phone, it's much harder to make an impact in the PC market because Apple's share is relatively limited. You know, closer to ten to fifteen percent. And so, but they probably, we're going to see a couple of points increase there, and we're going to see a couple of points increase in the phone, and this is going to be going down in history, the twenty twenty-six, twenty twenty-seven time period we'll probably see a substantial increase in market share for Apple. We'll look back upon this year and ask what happened for that change to occur. And that's why I think now, this last few weeks we've had this WWDC with AI. And we've had the memory crisis, which also is caused by AI, you know, restructuring the phone market. Some analysts are suggesting we're not going to see a recovery to 2025 levels of phone sales for three to four years. That's three to four years for Apple to continually beat down their competitors. And Ternus is going to take over just at this moment when there's probably this market shift happening. I think we might, and I've been sort of pulling punches on this, but we might be looking at Apple doubling down on a market share strategy that will see them shoot up significantly. Something that we've been thinking for 20 years they might do, but they haven't done yet. They've always been kind of happy with their limited market share, but obviously extremely high revenue share and extremely high profit share, to the point, by the way, where their profit share might end up being nearly a hundred percent, if not actually one hundred percent, going into the end of this year. So this is why I think the watershed is maybe AI, but we don't yet see the impact of AI on the bottom line. But we're seeing evidence that the market share game is changing. Because they made this decision probably two years ago to split into a fall lineup and then a spring lineup. Yeah. I think the only thing is that it's compounded by the memory crisis. That's affecting, by the way, it's, let me clarify, it's less of an effect on Apple because they manage their supply chain very well. They're also not squandering a lot of this memory on cheap product with no margin. But it's also not affecting Samsung as much because Samsung also has more of a mix of premium product. They sell lots of cheap, but they can always retreat back into the premium, and they then drop a lot of low-end product and they can still be in the business of selling phones. But Samsung is doing very well selling memory. They're far more profitable selling components than they are selling phones. So for Samsung to actually exit the phone business would probably be a blessing, because they'll be like, well, thank God we don't have to deal with that low margin, typically negative actually on many of their phones. That's interesting. But it's all the other guys, the Vivo, the Oppo, which are, by the way, numbers three and four in volumes. Nobody's heard of them because they're China mainly, and they sell in Africa and they sell in Middle East and they sell in parts of Asia. But globally those are the four, the two big brands. There's some other one, oh, in India I forgot to mention, and there's a lot going into these markets that is at two hundred dollars price points. And those $200 price points are no longer sustainable with the cost of memory that's out there. So what we'll see is contraction, Apple's share goes up. Premium, what they call premiumization, premiumization, I don't even know how to pronounce this, but the idea of portfolios going to the high end even if you're not an iPhone, that's been happening for a while. So maybe we'll look back upon this year again as something of a, I would, you know, market discontinuity, cracks, it's something cataclysmic, it's something substantial.

Judd: Well, it could be. It could be. I mean, it's interesting to me in that, I think what I can't get away from, and I'm curious if they care about this at all, but just because something is expensive doesn't mean it's special. And I don't know what, what do you think they think is special about the iPhone at this point? I mean, obviously it's the best phone.

Horace: But this is a syndrome that occurs every year. But people are always remarking about the very, very high end of the market. So the premium iPhone, when it reached one terabyte and it reached a thousand dollars, people freaked out and said, oh my god, the iPhone is so expensive, why is the iPhone, they're just gouging, da-da-da. Well, again, that's the top of that staircase. It is the highest of many products, but if you look down into the middle of that distribution, of that range, that's where the most popular phones still sell. They're selling at, you know, six to eight hundred dollar price points. This is the bread and butter iPhone, whatever is the right analogy. That is a mainstream product that's not going away. There's also, by the way, low-end versions, and there's last year's versions that are still for sale. Those end up being the cheapest. And so all of those things are gonna carry on. What they're separating is that the launch window that used to be fall for Christmas purchase, which, by the way, Christmas is gifting. So when you gift people, you want to always show maybe a little bit more of your generosity. So the gifting process typically also attracts the premium. And then the students and everybody else who needs to buy medium grade phones, they're still, but the 17 is going to be still for sale. It's done very well this year. It's gonna continue next year as the low-end product. The 18's coming out a couple of months later. No big deal. It's just that from a marketing point of view, from a ramp, the production point of view, they probably were seeing this already, that the mainstream phone and the low end phones were doing better later in the year, and that is later in the next year versus the end of the year. So I don't think this is a giant change in how Apple thinks about premium. I think it's just a cadence question that some people said, which I find incorrect. They said, well, the seventeen is doing so well they're gonna keep it going a couple of months more. No, no, no. This was decided years ago, but they're going to start to launch in a kind of a two-phase process. And, by the way, I think it keeps the excitement of the iPhone moving into spring. Because the problem they've had is that people would see new iPhones in the fall and then there'd be nothing new and the competitors would take advantage. Point of fact, Samsung always launched six months out of phase with Apple. So if Apple had a big event in the fall, Samsung would launch six months later, literally, so that they would come in when Apple was, you know, being long in the tooth. So now Apple suddenly has fresh product coming also at the same, you know, not quite six months, but into that Samsung cycle, and it's going to drive excitement. And as I said, China, their holidays are different than ours. So I think there's some value to this staggered approach. You know, if I were to design it, in fact, from the very beginning I would have probably said, yeah, why don't you launch around the whole year so that people are always looking forward to something. Again, Steve was the one who said, no, we do once, simple, yada yada. But it was because of clarity, it was because of the simplicity, and it wasn't optimal in the case of the five, yeah.

Judd: Well, but it's also an event. It's also an event.

Horace: It's also an event.

Judd: And what they had been doing over the years is releasing new colors in the spring, but maybe that just wasn't enough. Anything else before we sign off?

Horace: Well, again, the AI thing is one of those things that we've been waiting for two years. But we're gonna probably have another ten years before it propagates to most of the people who are using Apple phones. So let me, you know, a lot of commentators that I think are smart about this have been saying, hey, it looks like AI will be first experienced by the majority of people through an iPhone. And that's very profound. It's not just that, yeah, using a chatbot is not in and of itself an experience. People, I mean, some people are addicted to it, some people are using it productively, some people are becoming mentally ill, whatever. Yes, but the majority of people, your mother, my grandmother, or, well, I don't have one, but, you know, these types of laggard users or mainstream users are probably going to experience AI through their iPhone.

Judd: Yeah.

Horace: And that's, you now, we're now seeing just what that experience might end up being. It's gonna be a Siri and it's gonna be apps. And as a final line, the economic model, everyone's been thinking that you pay for AI, you pay ChatGPT monthly, you pay Anthropic, or you pay whomever. What Apple's probably going to do is not charge, unless, maybe you have a new tier of iCloud accounts, that, you know, if you're a frequent user you might hit a limit and then you'll be invited to pay more. But also the developers are going to be probably charging extra for these types of AI functions, and so what that goes into is services, almost without any drama, without any change in business model. Suddenly Apple will be seeing an acceleration in services, which has been the growth story for forever now. And I'm pretty excited that, it's almost, it's almost, what's the word? It's almost anticlimactic. The financial model of AI is anticlimactic. It's there, there'll be lots of revenue, but it's not even mentioned by Apple. They never talked about it in WWDC. Fascinating.

Judd: Horace, this has been great. Thank you for taking the time today. Thanks to everybody listening. We will be back with you very soon on Asymmetric.

---

Horace speaks carefully and with knowledge about opinions, facts and fantasies, and shares what he thinks John Ternus’s most important job is.

Topics include:

• Is a crash coming? 
• The CapEx problem
• AI and Apple
• Convergence vs. fragmentation
• Ternus and Apple's culture

Now in Beta for asymco one subscribers

Listen to Asymetric in your favorite podcast app.

Audio Recording

[asymco_podcast slug="asymetric" id="asy-137"]

---

Transcript

Judd • 00:00
Tim Cook got to go to China on Air Force One. Horace gets to talk to me. This is Asymmetric, Horace. There's a lot to talk about. You recently wrote that something terrible might be coming. I do not normally think of you as Chicken Little, but is in fact the sky falling, Horace? Are we about to see a crash?
Horace • 00:23
I didn't suggest it was happening, but I said, are we facing a crash? "Is there a crash coming", I think was the title. And this is a topic that's just been a worry for a long time, because we've seen the surge in the market value of so many companies.

But as I said many months ago, on a PE basis, it doesn't look so bad. But then we should ask, is that basis not valid? We should be looking at another basis, perhaps, because the PE ratio is not the whole picture. And so I'm putting out a series of posts. Because you see, the problem is this. There are some topics that are, let's say, not simple, and you feel like you should dive in, but you feel like you might leave the audience behind. And my struggle is to make it simple so that everybody can follow along, but without making it simplistic.
Judd • 01:32
Heavy lies the crown.
Horace • 01:34
Well, it's not that hard. It's just that sometimes you need to do five steps instead of just one. And you see this process online, where people are trying to fit everything into one tweet, or even now a thread of tweets — and there's a lot of long-form stuff actually going on social media nowadays. But I don't think this topic requires a very long single post. I think it just requires a few baby steps strung together. So I don't want to necessarily — well, we could dive into it right now, but I'll summarize at least.
So the problem with the PE is that it doesn't, first of all — remember, E is our friend. E is earnings. But it's also true that earnings are an opinion. Remember the old adage: earnings are an opinion, cash flow is a fact. And arguably I added to that by saying a share price is a fantasy. Literally, I mean it is a dream. Maybe I should say dream, not fantasy, because it could be a bad dream as well. But the fact is that earnings fundamentally are an opinion. They are a result of an accounting practice, an accounting convention — the so-called generally accepted accounting practice, GAAP.
And what it means is you don't put into earnings things like capital expenditures, you don't put into earnings the fact that you have a huge amount of cash. So things to do with cash flow don't go there, and things to do with balance sheet don't go there. And both are important in evaluating the value of a company. E comes from the income statement, and it does not cover issues that are outside the income statement.
And this is why a better measure — again, not a complete measure, but a better measure — is free cash flow, or FCF. And this is why years and years ago I said, because at the time the problem was people weren't valuing what was on the balance sheet of Apple, which was cash, huge amounts of cash — and so for that reason, even then, I suggested don't use PE, use free cash flow. Which, by the way, is the denominator, where the numerator is EV, or enterprise value, which includes debt and cash, or rather balances those out. So you divide — instead of doing the P over E, which is market capitalization over the earnings per share, you do it with the top being the EV or enterprise value, where you cancel out the debt, and then divide it by free cash flow. So that's as if you took not the market divided by the income statement, but you take the balance sheet divided by the cash flow statement, which is very interesting from an accounting point of view.
So anyway, I don't want to belabor this, but what I'm going to do is explain in writing, and using examples — especially the so-called Mag 7, or even expanded to the top ten most valuable companies.
Judd • 04:51
I look forward to that. My question for you right now, since we're here together chatting, is: you've long held that a PE of thirty-five is pretty much where Apple should be. It's fine. So in terms of free cash flow, what should that ratio be? What should that number be?
Horace • 05:13
Well, they should be roughly the same. So here's the beauty of this: the EV over FCF —
Judd • 05:17
So EV over FCF is 35.
Horace • 05:20
If they're the same, so — I put it this way: if P over E equals EV over FCF, it means that there's no weird stuff on the balance sheet, there's no weird stuff going on with the cash flow. It's all pretty much — the income statement is the best measure of the business. And in those situations you're not playing games elsewhere. Or — well, not that they're games. It's just that there's nothing invisible to you when you're looking at the income statement. And that's a nice way of analyzing a company.
The problem is today, with all the hugest companies in the world, which make up a disproportionate portion of the S&P 500 and a disproportionate part of the US economy and the growth and everything that's indexed to that — those few companies are out of whack. Their PEs are not at all equal to EV over FCF. In other words, there's stuff either on the balance sheet, or there's stuff going on in cash flow, which makes that PE ratio irrelevant. And this is what I have to expose. With Apple in the middle of it, showing how Apple actually conforms more or less to, let's say, an income-statement-driven view of its value, whereas the others are much more skewed, because so much of the money that's coming in is being spent on capex. Which means they are no longer — here's the catchphrase — they are no longer considered software or services companies. We're talking Microsoft, Google, Meta, Amazon. These are intellectual machines. They deal in software, they deal in non-tangible items. They deal with data. And yet they're investing today as if they're chimneys, boilers, forges, hammers, and our industrial-might giants, which require huge capex. There's nothing wrong with that. Sorry — there's a motorcycle. So the problem is —
Judd • 07:40
But haven't Google and Facebook for a really long time — I mean, they've had their own data centers, they've spent a lot of money. This is much more than it's been, but this is not entirely new.
Horace • 07:51
Well, it isn't, but the thing is it's way, way bigger than anyone anticipated. We're talking, on a combined level, at least half a trillion dollars being spent. It's moving the needle on everything in the whole economy. Caterpillar is having an amazing year. Energy companies are having an amazing year. We're bringing back nuclear reactors. We're bringing back water issues and energy issues, because these data centers are monsters.
So what I'm saying is that if these companies that historically have been seen as income-statement businesses are now suddenly balance-sheet and cash-flow businesses, we cannot evaluate them strictly on the PE basis. Apple is the outlier. So Apple is the exception that proves the rule. Apple is still extremely consistent with respect to all these ratios, how it used to be. But all the others have gone way off.
And by the way — I wrote a short note, a note of Bennett, which said basically that Google's cash flow is going to zero. Well, so is everyone else's. And what happens when they don't have even more cash flow is they're going to start borrowing. Which means not only are they destroying their cash flow, but they're going to be destroying their balance sheet as well, by putting a lot of debt in, in order to pay for these infinite data centers. So there's an interesting dynamic.
I would say we're in uncharted waters, but we have to then put a little bit more effort into trying to figure out whether we're in a bubble or not, because this is the thing: all of that could be supported by enormous growth in revenue, enormous growth in profitability — and there is growth, tremendous growth. It's 50, 60, 70, 90% growth. But their investments are growing even faster. So this is why some people are grasping for ways to detect it. I've seen half a dozen indicators out there — everything from the Warren Buffett indicator, this or that, that things are out of whack. But all these indicators were dependent on the ratios of the past. We are in a situation — it's not like the dot-com era, it's not like the Great Recession, it's not like the Great Depression, it's not like all these previous bubbles and crashes. If you applied those metrics, many of them may be true today, but then you have other things that are actually even more important. So that's what the series is all about: we're going to try to get to the bottom of it. And I will propose at the end, this is the best way to think about a business.
Judd • 10:35
What's your best guess right now?
Horace • 10:37
So the way to look at it is: given the growth projected forward, how many years will it take you, as an investor, to get your money back? If you put down a hundred bucks on a share, or a portion of a share — let's say. I don't think anything's a hundred bucks anymore. So you buy a share in Google and —
Judd • 10:56
How long's it gonna take?
Horace • 10:59
Sorry?
Judd • 10:59
A pair of pants is a hundred bucks.
Horace • 11:01
Oh, well, you've gotta get your Dockers. But a hundred bucks invested today has a payback period based on cash flow and growth. So those are the two most important metrics. And the problem with the growth is: are you going to look at the last quarter and project it for the next five years? Are you going to look at the last year and project it for the next three years? So the whole thing depends on how much of that growth is sustainable. So all of this comes down to — okay, well, we don't know for sure, but let's play with some scenarios. And then you get a good idea: am I going to wait five years, ten years, twenty years?
And in the case of Tesla, by the way, which is also among this group — it's an outlier. Its PE is 400. Even if it grew at 70%, it's not going to get back. And of course the sad truth of Tesla is it's actually contracting as a business. So it's actually the only one that is going negative in growth. And yet it's also the only one that's accelerating in PE. And so there's a huge anomaly. There's a huge elephant in the room — we talk about theory, and then it demonstrates — yeah, well, there's a robotic elephant. Or space robots. There's so much mystery, because that demonstrates also how sentiment, optimism, euphoria, and memes all play a part. So anyway — sorry, this is a bit, as they say, in the weeds. But I feel it needs to be done.
Judd • 12:47
Yeah, well, let's put a pin in this. We look forward to seeing what you come up with. Please, as always, show your work. And we'll be paying close attention to your follow-ups to this. It's hard to not talk about AI. Would it be fair to say that services — which is becoming more and more the centerpiece of Apple's business — would it be fair to say that services is the most vulnerable part of Apple to AI, in the sense of — because services boils down to —
Horace • 13:32
The US? Yeah, I get your question. Well, it's a great question, because I think there are a lot of hypotheses out there about how Apple could be disrupted by AI — and I use that term carefully and with knowledge. The scenarios which are disruptive to Apple would be not just for services but also the hardware itself: that the business would move away from what the user interfaces today for the consumer.
But you don't actually think that. I do not. In fact, I think I can set that out. The underlying thesis is that we are visual machines, or visual animals. And as long as vision is involved, I think Apple's going to do very well. If you say, well, it's audio in the future, or something that is much more in the background, not the foreground — it's ambient. Computing becomes ambient, then there's no interface. If there's no interface, Apple has a hard time. But I just don't believe that future is likely.
And we — I think we mentioned this before — the services aspect: now people are starting to wake up and say, wait a minute, actually services could get a big sustaining boost from AI. Just days ago there's been an idea put forward that Apple could launch an agent store. The way there's an app store, there could be an agent store. And what an agent is, as you would imagine, is like an entity that you hire to do things for you. And it's very possible that Apple could embody or integrate agents into its services, and that the discovery of agents, payments — all of the things that are related to the economics of AI — would go through Apple. And people are finally realizing that, yeah, as I've been saying for three years, distribution is just as important in the age of AI as it was in the age of software, as it was in the age of computers, or any other pieces of equipment that we have ever bought.
So agents are possibly the embodiment of AI that are actually consumer-oriented. I hate to go back and forth on this, but the fact is the AI business is still mostly, right now, very few professionals are using it, and quite a few businesses are using it. But it's not yet consumer-driven. Maybe chatbots are, or entertainment, maybe companionship — but as I see it, no one's quite gotten the killer app. And that will probably be discovered through introduction via an iPhone. That's the platform for discovery and everything else.
So I think many people now are starting to write — and by many I mean the analysts on Wall Street are starting to write reports suggesting, wow, AI could actually be very sustaining to Apple, to the point where it's going to ride the wave and be the one capturing a significant portion of the profits. I would even say it may even capture the majority of the profits, because it's at the intersection of the consumer and the device. And that's not going to change. And I think the attacks by OpenAI, to try to replace that device with whatever wearable Johnny Ive and company come up with, will fail. And I quote Johnny on this matter. Johnny Ive said, "I'm nothing outside Apple." He, and every other person who's left Apple who's smart enough, has said the same thing. And yet here we are with this idea that, oh, I'm just going to sell my services to an AI company. Anyway.
Judd • 17:48
And what it feels like he might be doing is something that — I mean, it's touch-based, it's voice-based, and all those things are much slower than a visual interface. You also recently wrote about using the Apple Vision Pro outdoors, which I'm skeptical of, at least for me personally. I would not want to do that. Maybe I'll have no choice. But it's not something that is —
Horace • 18:20
It's not going to be the device as you see it today. So the puzzle for me is — we have rumors now that Apple is doing glasses. Generally I don't comment on rumors, but it's obvious to me that there would be a wearable. What about a monocle? Yeah. Or — what is it called? A pince-nez? Yeah. But the problem with these devices is that maybe they're mostly — what I would say is — input devices, like cameras or other sensors reading things. There are also rumors of AirPods having cameras. The view of the future I have is that there will be wearables which have augmented reality and you could wear them all the time. That's not the case of the Vision Pro today. But I think if I were to imagine a future, theoretically it's possible to do that, and that would be fantastic.
Anyway, back to the threats at hand. I think AI, for me — at least the version we have, which is not yet AGI — it's a nice extension. It helps with a lot of tasks, it improves some productivity, decreases other things. But it isn't disruptive yet. It might be disruptive to some type of work out there. So it's disruptive to people. But none of the players — okay, I take it back. There's a couple of companies out there that might feel it, and are being punished right now because of AI. One is Adobe; another would be so-called SaaS, or Software as a Service. You have companies essentially that are providing tools or tool chains, and those might be very easily replaced with the person who does those things but happens to be an artificial person. So if there's not that much value added with those technologies, then yes, they would be disrupted. But even they, I think, will bounce back.
One of the issues is — especially with software — and I don't see a lot of conversation about this — people talk about vibe coding, and writing code very quickly and being very productive with programming tasks, but that doesn't take into consideration the maintenance of that. And the maintenance, as anyone who does software engineering knows, is not so much about how much code you write, it's how well that code can be managed, maintained, upgraded, and everything else. And that's the engineering part of software engineering. When something other than you writes something — even if it's a person — you're going to have to deal with that.
This was a big problem when I was a student: you were told, comment your code for your own sake, so that when you go back to it you know what you did. But also, if you have to hand it off to somebody, it's more self-explanatory. And it was a very bad practice — almost an antisocial practice — to have cryptic code that nobody can read, therefore nobody can maintain, and you're just doing it to preserve your own job, so nobody can fire you, because nobody can fix whatever you wrote. That's just not considered good practice. And now do we have the equivalent of machines writing code, and they don't even know what they're doing? I mean, they're just basically copy-pasting a lot of things. And so they become spaghetti code, it becomes intractable. You can't understand what they did. And so it becomes unsupportable, unmaintainable, and that might cost you even more to rectify than if you had a human write it in the first place. So either machines get far better at creating code that is inherently supportable and they can explain themselves — but there's so many things wrong with this picture as far as the long-term things, right?
Somebody wrote, saying, yeah, it may look cheap to write code, but you're going to get the invoice on maintenance years down the road, and you'll realize it wasn't cheap at all. So there's a lot of things about that, that I don't see disruption happening here. But there are some things. A lot of work, like legal, tax, or accounting — a lot of that — I've used it in these instances where it's a better search. A lot better to search with Gemini, for example, than it was with plain old Google. And I'm learning a lot of things. Although I'm a little bit cautious, because sometimes, as it says as a disclaimer at the bottom of every page, this is not always accurate — which is a worry sometimes. So anyway.
Judd • 23:38
You know, Google wasn't always accurate either. So in all cases, buyer beware. You were saying something a minute ago in terms of having an agent store, and I'm not sure if you are a reader of Neal Stephenson, as I am. I don't know if I'm current — in fact, I know I'm not current — but a few years back he wrote a book called Fall, which was a follow-up to a book called Reamde. Fall takes place in an age where AI is very common, and this is the thrust of the story. But it's very common, and every family, if not every individual member of that family, has their own agent — to do things, process information, give them answers, tell them what to do, et cetera. Which to me leads to the ultimate form of reality being fragmented. So everyone is living in a completely different world based on their agent. Should there be a required agent? Like, if you live in Helsinki, should everyone have at least the Helsinki agent, so they know just the basics — so there's a basic frame of reference?
Horace • 25:01
Yeah, but it's a huge problem, because we don't have a lot of shared experiences anymore, right? Even within the same family, people are not consuming the same content. It used to be at least there was one newspaper in the house, there's one TV set, and more or less everybody watched the same TV. And we ended up with a situation where, even when there's a common space, where people live together, people are looking at their own screens, and they're looking at a different reality. And then if that is intermediated by an agent who is essentially, let's say — rather than calling it an agent, I would say, let's say it was a personal assistant. And the personal assistant is like a gatekeeper. Chief of staff, whatever. And they are intervening even in the question of whom you meet, and what you choose to do. They kind of act on your behalf. So there's a lot of problems with that, because it's not just what you consume, it's also acting on your behalf in all aspects of life. It's kind of a little scary, if I'm honest.
But let me put it through a different perspective, which is that already there's a lot of difference between people from different countries. It's fascinating to me, having lived in many countries, how different the perspectives are on very basic things. In many ways, by the way, we are also getting together more. And let me explain what I mean. The globalization phenomenon has meant that people are more aware of, let's say — Americans are more aware of the World Cup, for example, soccer, which was something nobody cared about. Europeans are more aware of American football, and they're sort of kind of aware of baseball. And at the same time, you have shared devices like Apple, the iPhone, you have shared brands — everybody knows about Nike. Americans will know about European cars, and vice versa. So at the commercial level, because businesses have grown to be global, products and brands have kind of become uniform. It used to be that you went to a different country and you couldn't see on the shelf anything you could recognize. They ate totally different foods, they had totally different tastes. But now you can kind of agree that, as far as packaged goods are concerned, there's commonality. And that has a lot to do with the globalization of business, right? The fact that I can get a banana in the middle of Finland — even though it's a tropical fruit — and the banana I get is as fresh as it is in Costa Rica. That's an incredible story.
Judd • 28:24
Pretty close.
Horace • 28:25
Well, it's green. I think — and by the way, all these bananas are genetically identical, because they're all clones, because they're all made from the same exact genetic material. And if that gets wiped out, then we have to reinvent bananas, if you will. So there's a lot of funny things that happen because we have these modes of transport — these ships and airplanes around the world that constantly tie us together. So we have that. That's a positive. But on the information side of things —
Judd • 29:01
So we've got the same bananas, but completely different sources of news and information.
Horace • 29:05
Right — the same bananas, but at the same time — oh, I'll give you another stupid example. Americans have baseball caps. Well, baseball caps are global now. But I can tell you, when I grew up in the 70s and 80s, nobody outside the United States wore baseball caps. Maybe Japan, because Japan, weirdly, was colonized by baseball. And yet this style of headwear was never seen. Everybody had different hats, different styles. You had berets, and you had fedoras, or bowler hats. And yet now there's so many things that are uniform — in some ways, boring as a result. We all kind of wear sneakers. We all look the same. We all dress the same. Probably the music is also very common.
And the other thing — I'll give an example. Rap music. American music back in the 70s, 80s was well known because there was a global music industry. But people in Europe did not necessarily make a French version or German version of what would be a pop song. Let me put it the other way. ABBA was European. And ABBA wasn't all that resonant in the US — although it was enjoyed, it wasn't like our thing. People were much more into either rock and roll, or rhythm and blues, or Motown — American.
Judd • 30:49
Well, America had a disco phase.
Horace • 30:52
Sure, sure, but — there was, and it mostly came from Europe. But you also had indigenous American music that was enjoyed in Europe, but it wasn't felt that it was something that they produced themselves in the same way. You can do Motown in Europe, okay. So the thing is, though, today rap is universal, and it's from the Philippines to South Africa to Japan. That genre has caught on with everybody.
Judd • 31:23
Who is your favorite Filipino rap group these days, Horace?
Horace • 31:25
I have none, I'm sorry. I'm not aware of the individual acts. But what I'm saying is, I was observing it with youth behaviors — that they're acting very American here in Finland. At the same time, though, they don't understand a lot of American decision-making. They have no perception of student debt here. It's such a strange idea. Or medical bankruptcy.
Judd • 31:54
Yeah.
Horace • 31:56
The bottom line on this is that we can hear these stories about fragmentation, disassociation — but at the same time, there's a lot going on with unification of experiences. And most of all, think about your iPhone. Everybody in the world speaks iPhone. Everybody in the world recognizes and interacts with a computer the same way. And even before the iPhone, that was the case for computers in general, whether it was Microsoft or Mac. We — the Americans — invented how to compute effectively. And everything else in the world just completely disappeared. It wasn't that they didn't have it, they just disappeared. The French way of doing things, or even the Soviet Union had its own computers. But it's one of those things that we don't pay attention to very much, because — well, it's —
Judd • 32:49
It's too obvious. What exactly do you mean by that when you say the American way to compute? Do you mean — because Apple and Microsoft dominate personal computing and they're American companies — or is there something —
Horace • 33:02
First of all, the American computer is based on — I'll give you a very simple example — the byte. The byte is eight bits. Eight bits is enough to encode ASCII, which is the code for characters, and then you have the alphabet and the numerals. And the problem is, if Japan — let's have an alternate history. Japan wins World War II — Japan and China, maybe. Imagine that Japan and China together, with their kanji alphabet, would have come to dominate Asia, and then more and more of the world. Or Russia with Cyrillic — but Cyrillic is still a limited character set. The problem is you have fifty thousand kanji characters. You wouldn't be able to fit them in one byte. You would have to have a computer architecture that was bigger words, as they say. The layout on screen also — what happened with American characters on a screen is that you have eighty characters on the original terminals, eighty characters on an IBM punch card, and that was a line. And from that came the screen as we know it. Now, the form factor of the screen, the page layouts, the metaphors of Microsoft when it copied Xerox, and the Macintosh, the file folders — everything to do with layout on the screen was what an office looked like. All these things are, if not American, Western-centric. And the American keyboard is not the same, by the way. French keyboards are different. So you have all these weird things that were introduced by —
Judd • 34:56
There's a croissant button.
Horace • 35:00
The fact you have all these accents. But the thing is, the world adapted and internationalized what was effectively an American worldview, or compute view. And now, by the way, in AI it's the same thing, although it's English-led. The techniques can work in any language. But the leadership that comes from the United States, as far as AI is concerned, is — more or less people aren't questioning it. Who knows how it would have been different had it come from somewhere else? But I think the aspect of unified compute is not necessarily a counterpoint to the disunified cultures. In many ways we are much more alike culturally than we used to be. The Arabs in Syria — oh, to give you another example. If you go to Saudi Arabia — a friend of mine went there recently; I haven't been. I've been to Dubai, but I'm embarrassed to say I haven't been to Saudi Arabia. But if you go to places in the Middle East, it's burger joints. It's chicken joints. It's fries. It's ketchup. These are not what Bedouins ate. And probably there are restaurants that still serve indigenous cuisine, but they are one in a thousand, literally. So it's amazing to me how that American culture is everywhere. Maybe the Middle East is more skewed than, let's say, India. But still, the Indians know America very, very well indeed.
Judd • 36:50
Not that I wouldn't love a traditional Bedouin feast, but a hamburger joint works pretty well.
Horace • 37:00
It's exactly right. The economic models that scale, and that are replicable because of their inherent efficiencies and demonstrated value, are quickly adopted. What seems to be divisive, I would say, runs not on cultural lines, it runs much more on human nature being roughly divisible into several personality types. And then those divisions are amplified by the media you consume, and possibly by the agents in the future. And so you might be predisposed to lean one way or another politically — which, by the way, I think is one of these unspoken truths, that somehow there is a genetic component to political affiliation. And it's very difficult to actually get people to switch. Once they establish their — of course, there are the independents who claim that they switch, and probably independents are the majority — but nonetheless, there is — I think what algorithms discover, because they offer you things, and you react to things, and they offer you more of the things that you tend to dwell on — they immediately are able to characterize you based on your consumption. And then they create a dysfunction, because they then feed you things that exclude what other people see, and you become radicalized.
And you can do this on very basic things, like cats and dogs. Cat lovers never see dog pictures, and then they grow up and say, oh, I hate all dogs. I'm exaggerating, but I think this idea of, we end up being divided — we're unified with the fact that we use the same computers, the same interfaces, possibly even speak the same language. A lot of people speak English. But then you divide yourselves — you divide into cat people and dog people. And the more you consume, the more you are the radicalized cat person. But do we want to talk about John Ternus at all?
Judd • 39:25
Horace, who would win in a fight at their primes: Steve Ballmer on stage at a developers conference, or John Ternus preparing for a WWDC demo video, in a skin-tight t-shirt?
Horace • 39:42
Easy. Ternus is going to completely crush Ballmer. Let me — oh, yeah. Look, Ternus was —
Judd • 39:52
Ballmer is hyped up on all those sales figures and the developers —
Horace • 39:55
All that — pulls a muscle, screaming, "Developers, developers, developers." He pulled a muscle on stage. Ternus — was he what, varsity in college? Swimming? The guy can take his team off and do hikes in the mountains. Swimming — I did swimming in high school. And swimming was —
Judd • 40:17
That might help you if you're being surrounded by salmon, but —
Horace • 40:22
We're talking about dolphins. No, dolphins can be very powerful. But look, swimmers — come on. Well, if anything, he can run away from him.
Judd • 40:38
He'll need to run. I mean, he might have more endurance — that would be the thing. So I would say Ballmer has more brute strength, and Ternus has more stamina. So if he can evade for a little bit —
Horace • 40:56
Could he do one of those kick turns in the water? I don't know. I'm a lover, not a fighter. I don't know how to evaluate fighting ability.
Judd • 41:09
But I will say this about Ternus — I don't know if you're aware of this, but the company he worked at before Apple, which is, I guess, thirty-some years ago at this point, was a VR company.
Horace • 41:21
Yeah. But that was his first job, and he was there only a couple of years. And I've read —
Judd • 41:26
Isn't that exciting for all of you Apple Vision Pro aficionados?
Horace • 41:29
Yeah, yeah. No, I think he knows that. He's got ideas there. He did something in grad school or undergrad to do with helping people who had some sort of disability — assistance, robotics or something, I forget. And then he went to a VR company, and that would have been like 25, 30 years ago, like you said. And then he went to Apple and has been there since. I think the first product he worked on was a cinema display, of which I have one, by the way — or had one, anyway. Tremendous product.
But the thing I said about him is that I use the word engineer. He's like a real engineer, and he's an engineer who can connect with non-engineers — but "engineers" is too broad a topic. You can say, well, he's product, or he's R&D — but he's all of those things. I think it's important to get a person like that to retain talent. It's not just motivational. There's a big problem, I think, Apple has of turnover. Now, it's no bigger than any other company. But in the case of Apple, they need to keep people more than other companies. They need to have that loyalty, because things take a long time. So if the project you're on isn't going to see the light of day for seven years, it would be a pity if you have to switch over the team three times during those seven years. There's going to be big problems with continuity if there's turnover in product teams that are committed to long-term projects. So that's an issue that Apple has to contend with.
And the way Steve dealt with it — I think this is where the questions of culture come in — I think Steve offered to his greatest people the chance to do their best work, and not to be distracted by politics, by management which has turf wars, or has an agenda. Managers have personal agendas, empire builders. And once that work is hopefully finished, then millions of people will use your product. I think these are the motivations of the engineer, of the creative and analytical type. And by the way, you could argue the same thing for someone who makes a movie. If you say to somebody, hey, come and work on my project, here's a bunch of money — but if you're working with this great team, and we have a great project, a great script, a great director, that movie could end up being seen by a billion people — you might say, you know, I'm going to give it my best. And forevermore I can reflect on that. And whatever I do in the future, the glory of that product will shine upon me forever. And that doesn't mean necessarily you get paid a lot because of it — probably not; probably you'll get a little bit of residual value from it. But you'll have that on your resume. And believe me, people are going to come out from everywhere asking to work with you, simply because you're glowing from the glory of that project.
So I think — now, that sounds vain, but it's more than that, because I think the engineer devotes so much of their time, their life, early in their career, their education, their learning, and everything else — it's brutal. It's quite a challenging program to get through. And at the end you feel like, hey, I'm not going to use 10% of that capability — I hope not, otherwise I wasted all this effort. Secondly, I hope that I can make an impact, not just make money. And you can argue the same for a doctor. A doctor probably feels great if they get a nice job — maybe they're a lobbyist, maybe they work in compliance, maybe they end up in something else with their medical degree. But if you're a practitioner, and you end up saving people's lives — or, in the case of our friend Dr. Tommy Korn, he saved 23,000 patients' eyesight — I can't imagine how satisfying that is. Same thing for the creators of the Macintosh, right? And they say that's why he had them sign the inside of the case. You guys really made a miracle, and those people can be proud.
So that's why I think Ternus — or JT, let's call him JT — I think he will offer this to the people working there, as did Steve, and as did Tim. But in the case of Tim, it was a secondary thing — no, that's the wrong word. It was through other people that he inspired, right? He would say, well, we have good people, and they'll inspire you. So I think that's the directness of JT's connection with the teams — that he was in the trenches with them, working on the same things, that he stuck it through 25 years, and now, as a result, he really made it big. Not that they're all going to be wanting to climb the ladder — that's also not necessary. But anyway, motivation and retention.
Remember, we talk about satisfaction a lot. The satisfied customer is number one priority. Well, the satisfied employee is probably the number two priority. Satisfied employees at Apple are a really, really high priority. And I remember being asked years ago, what do you think keeps Tim Cook up at night? And I said: employee retention. It was true then, it's true now, and I think it's going to be true for JT as well.
Judd • 48:10
So Steve Ballmer is six-five. I could not find a published height for Ternus, but he looks to be about the same height as Tim Cook, who's about six feet. So I think one-on-one, if it was like a bar fight, I think one-on-one Ballmer would have an advantage. But basically what you're saying is, more people in the bar would back up JT than Ballmer. Is that —
Horace • 48:39
Well, yeah. And I think JT's going to —
Judd • 48:41
He would inspire more people.
Horace • 48:43
He's never going to be within reach. If Ballmer's got the reach, with the height, I think JT's going to dance around him and get him with a lateral move, whatever the term is.
Judd • 49:04
A hook. Sure. Yeah. Horace, this has been fun. Thank you for taking the time. Thank you to everybody listening. We will be back again very soon on Asymmetric.

For the first time, we will be bringing all Asymco Podcasts under the same roof, included in Asymco One subscription.

Asymco One subscriptions will include access not only to the new episodes of the Asymetric Podcast, but also to all classic Asymco Podcast content.

We have already started the process of remastering, transcription, and indexing archives of: The Critical Path, Asymcar, and Asymetric.

Watch Asymco Podcasts for the latest updates.

Learn more about Asymco One.