Ben Thompson: Apple still doesn't get the Vision Pro — just put a camera courtside and stop editing

getting paid.

Okay. Well, we have Ben Thompson in the Reream waiting room. Let's bring him into the TV at Ultra.

Fantastic. Ben, how are you doing?

Here he is.

Good.

Thanks so much for hopping [laughter] on the show in such short notice. Uh I loved your piece.

I have I have takes to drop. So I am I am

Let's go.

Let's go. Okay. So uh I mean uh drop your first takes and then I want to know uh is there anything to what I was saying that that Apple actually does read your work and they do want to do it but they just can't because of some contract.

Oh I thought you were talking about watches.

Oh no. I was talking about watches. What's on the wrist man? [laughter]

Uh I have a tutor GMT.

Very basic.

But highly recommend it. T's favorite place to

start. And and it's covered in diamonds, I assume, like the rocks.

Uh it is not. [laughter] It is very basic with a rubber band cuz rubber bands rule.

Oo. Okay. Rubber band was good. Yeah.

So, uh and I disagree with your take,

please.

Um so, my overall view, NBA was on the Vision Pro on Friday.

Yeah.

They had an NBA clip when they launched the Vision Pro. So, I was there when they watched it. So, I got to try it that day. They took that clip out for the demo that shipped and that was in Apple stores. So you only saw that clip if you were there the the first day. And so I was a right to choose which I maybe lends itself to your argument or whatever it might be.

But it that was for sure it was like 3 seconds. It was the clip that sold me on the Vision Pro more than anything cuz you felt like you were there. It was amazing. Yeah.

So they have this finally have a live sports thing which is a big deal. They demonstrated that they can show stuff live and it worked. It worked well. I don't think there was any technical issues. Uh the cameras are quite small now compared to what they used to be. You could see them on like the sideline table and underneath the basket.

So this technology works. You can watch things live in the vision pro and it should be amazing and it sucked [laughter] and and it sucked because you're getting jerked around all the time. You're not you're there. You feel like you're there. It is immersive. The unemersive part is the Apple part, which is some producer in a truck is moving you around. There's a pregame show. I don't need a pregame show. I'm sitting courtside. I can watch the players warm up. And the reason why I disagree with your take that Apple knows this. Okay.

And wants to do it. Is that every Vision Pro video has the same problem.

Okay.

The Metallica video, super cool. You walk in. The opening scene of that is amazing. Well, there's the little bit where they're in like the the green room or whatever, but you're walking in, you're walking behind James Heatfield and you're in the crowd and it's like walking through and people are reaching and it's like it's so amazing and then suddenly he's going up the stairs and boom, you're jerked to somewhere else and you're seeing him come up this on the stage. You go back, they cut this MLS video which they have the rights for. They're overpaying. You want to talk about the F1 deal? Look how much Apple's paying MLS. Like one of the most absurd overpays in the history of sports. I think they can do whatever they want here. They cut a I love laugh tracks, but it is [laughter] very disruptive to hear. Um

they cut like a season and review video. Yeah. That had like 56 cuts in it. It's like done by

What do you think it is? Do you think it's like Do you think it's production teams that are just trying to create work for themselves because they go in there and they're like, "Hey boss, actually the people just want to they just want to put their goggles on and hang out." It's like, "Well, then what are you doing here?"

I feel like there's just a lack of confidence here. like like what I it feels very like it feels when you say it it's like halfass that we're going to go out we're just going to set a camera there and you can sit there and watch the game. It's like no

there's this pressure in part because it's not popular. We It has to be a big production. We have to have a pregame show. We have to have dedicated announcers. You have to do all the things. The end result is you get six games which are physically uncomfortable to watch because you're getting jerked around and like they could have just the Vision Pro there. Yeah. For every game and I would pay a lot of money to watch that. They could have it at every concert and this sol it's what's amazing is it number one it would make it better. I'm fully convinced of this. Yeah. And number two, it would solve their content problem because you could suddenly have tickets to every single concert in the world, to every single game in the world. It's it is a classic tech issue where you put in the upfront cost, you buy the cameras, you install them, and you have marginal upside forever everywhere. and and it it's latching on to a major trend which is live which is this idea that when everything's commoditized online everything's digital live experiences are worth more and more and more like you go back to like the Aerys tour and Taylor Swift like a big topic of discussion at that point because like this is a something people are

sports specifically are like pretty immune to to AI because nobody wants to watch like the AI box play

no watch AI basketball right yeah it's the it's the human aspect And I don't think it's the thing that was most crazy to me is that it was like geolocked to Southern California.

I get so I get that. I I'm not going to like be on them for that because but part of it the reason why it was geolocked is cuz Spectrum spent all the money to do all this production

and so of course they get the benefit. It's in their catchment area. But you don't need to do all that. They're overthinking this. like literally set a camera on the sideline and do nothing else. I will be happy as a clam.

Okay. Okay. Uh it is expensive to do immersive production at the same time. Blackmagic has the Ursa Cinei immersive that have you seen this thing? Two fisheye lenses on the front 16K. It's it it cost $30,000. And it feels like if this is the case, there has to be an opportunity for someone who's a little more agile, you know, maybe it's Red Bull, maybe it's some league that's not front and center to just go and do this. They have to pay 30K up front.

What's expensive?

Yeah. 30K is nothing. Okay. Like we're we're all in the tech industry. We can laugh at figures like that. Okay. So So that's an upfront investment. Like the problem is they tacked on all these marginal costs to this production. So, when you logged on for the game, there was a dedicated pregame host and show for the Apple Vision Pro viewers.

That's almost certainly not going to be as good as the main pregame [laughter]

that where they're paying where they're paying like hundreds of millions of dollars

is sitting courtside and watching NBA players warm up and make 57 threes in a row cuz they're incredible. Like, I don't know if you guys have ever sat courtside, but like it's it's

it's amazing. Like, speaking of watches, this is where I spend my watch money, okay? It's it it's an unbelievable experience and and you and they have all the like they have like uh for example they have like replays in there

which okay yes you can see the better of replays. You know where else they show replays

on the scoreboard. [laughter] How else can I see the scoreboard? I can look up because I'm wearing a vision pro that's immersive and has this fisheye camera where I can see all around it. I don't need a scoreboard bug down at the bottom. I can look at the scoreboard. [laughter] Like I actually during the game while watching it, it was actually kind of hard to see the scoreboard bug cuz it was way down at the bottom. So I was looking at the scoreboard in the arena the whole time. Yeah,

that's the whole that's the whole thing. You're there. It's it's And so all that stuff, the the pregame show, they had dedicated play-by-play announcers and and analysts for the Vision Pro. They had these multiple cameras. They had a production truck who's choosing how can I upset Ben this time and every like [laughter] the other cameras. You didn't need any of that. And and and you $30,000 is a onetime cost. Yeah. You spend that money once and you put these cameras in every arena for every game everywhere and then you you suddenly have this exclusive selling tickets to live events in a way no one else can. I think it's a huge like it seems so clear to me. This is the Vision Pro use case. This is your portal to every live event in the world.

What do you What do you think about Meta Meta's efforts on uh NBA uh partnership as well?

Specifically with Oculus

the Vision the Vision Pro is better. I mean like the the like the Oculus has advantages relative to the Vision Pro. Actually I love Mark Zuckerberg's post and the Vision Pro came out. It's like oh we're not surprised. We could have done that but we didn't. Which was valid. Um I'm also not particularly interested in the use cases that are good for that. I don't care about gaming. I don't care about a lot of the other stuff. Yeah.

Uh, like this is for me the killer use case for sure. Like being able to

Yeah. Yeah. I mean, it seems like the next uh the next Quest, if they stay in VR, will be better. We talked to James Cameron uh and it seemed like he was pretty excited that they might have gotten the same screen from the Apple Vision Pro. like yeah, they're two years behind, but you bring that screen into the next Quest 4,

which is fine, but then there's trade-offs for that as far as like number of pixels and the the field of view and all those sorts of things, but like the the the whole point is that $3,500 feels overwhelming, like for what you get.

Mhm.

$3,500 to be able to attend any sporting event in person is one of the greatest deals of all time. Yeah, I used to be in Taiwan. I used to be in Taiwan. I would fly over for specific games like um and guess what? I would spend more than $35 on my ticket. I'm not flying in the back of the bus at this point. Sorry. Like so like like the this would this is

it is like it's like when the the Apple Watch when they realize oh people it's f it's fitness. Remember they watched with all these lists of different things and what you could like remember you could like draw your Apple Watch and it would like draw you [laughter]

the heartbeat texting to send someone a heartbeat that never went

who's the Apple who's the Apple exec that's the most hardcore NBA fan cuz I feel like I feel like you got

I so I I know how much Golden State tickets cost because I sat well I I actually I was poor so I was in the second row behind the seat that was inscribed with Eddie Q's name on it right there.

Uh yeah, he would he would he would be the he's the guy that needs to argue for this.

Okay. So, do you think this is a next season they might get it right? Are they already pre-baked on the next on the next because they have like four more games. It feels like they could take this they they have the footage like they could just not edit the next game, not do that broadcast and do exactly what you're prescribing.

I hope so. Look, free advice Apple. This is your That's why I made I I've written this. So number one, the reason I disagree with your take is because every video they make suffers from this problem. They're they are produced in a TV style for a device that is fundamentally different than TV. That's the core problem.

Is [laughter] there is there at least a a possibility that Metallica also said, "Hey, we don't want a substitutive effect. We don't want people."

Yeah. So it's so funny to think the same thing.

Lots of other videos. They made lots of other like every video they have suffers.

Yeah. Even the Alicia thing. Can you can you imagine if the if the NBA actually sold a ticket where like every 30 seconds you had to stand up and go watch from a different [laughter] from a different view? It's like they're like it's amazing. You see every angle. You're like, "Please, sir, I just want to sit and watch."

I turned it off. I I I I watched both I watched half the game on TV cuz it was it was uncomfortable.

What do you What do you think is the long-term like like viewing like ideal viewing experience? Is it you have it on TV and you have your your Apple vision and you're kind of like, you know, based on how closely you're paying attention, you're either like locked in

with the vision pro on or you have it on.

But the

I think that a big thing for sure, the the biggest missing experience is of course I want to be on my phone half the time like uh like any person.

I do have the the the M5 Vision Pro. It's the first time I ever took a demo unit. I've always like declined to do that, but I'm like there's no way on earth I'm paying for this, but I [laughter] do actually want to see if there's any difference. The my actually biggest takeaway is the pass through is much better. This might be I didn't do a I didn't do a direct compare. That was just sort of my perception. And to be honest, I haven't used a Vision Pro in a long time, so who knows how good my memory is. The pass through was really good. I had no issue using my phone at all. It did seem better than it used to be. Um so, but maybe that might be placebo just just to be clear. Getting screen time. Getting screen time through another screen [laughter] is elite.

It's a lot.

It's It's We got levels.

Wait. So, your screen time app could potentially show more than like [laughter] I had 26 hours.

26 hours today because I was in the Vision Pro using my phone, my laptop.

What a world. What a world. Um that's funny. Um well, uh I mean, are you optimistic about any other Apple Vision Pro like opportunities? Because, uh I was looking at this $30,000 immersive camera and I was thinking like I I was actually advising some friends who they shoot a much more uh like like evergreen podcast like these really definitive interviews and I was saying maybe you should start shoot they were shooting in 8K because they were like hey maybe one day we'll license these to Netflix we want the extra resolution we're not distributing in 8K but we have it now and I was saying well you want to go further why not 16K why not immersive maybe you should get this thing start recording now this particular device can only record for like 45 minutes I and then you have to swap the cards or whatever. But I was like maybe you should be doing the flow maybe not this year but maybe next year. Uh are is there is there some opportunity where someone

who's there's lots of opportunities like like and I think like there there's real enterprise opportunity like for like especially with the pass through and things that you can do along those lines. There's productivity sort of possibilities but you need like what is the anchor thing? What is it everyone knows that this does? And again I think the Apple Watch is a good example here. It took them a while to get to that. They launched with the physical fitness stuff, but it wasn't queer that that was the thing for a few years. And now every vision, every Apple Watch thing is fitness, fitness, fitness, fitness. It saves your life and you know, uh, all those sorts of things. And to me, this is it. It It is you get live in your living room. That's the vision pro pitch.

That's going to be fun. We'll just have to wait another five years for them to back down from their their opinions. It's but

we're going to find we're going to find out the day. You did write this two years ago. Like I wrote this the day I wrote this the day it launched. I said

what's amazing about this is you don't need production. All I want is to feel like I'm there and it delivers. It lets you do you know do you of course you don't know you're there. The resolution isn't perfect. There is a bit of tearing if you go super fast. Like the peripheral vision's not not not perfect. But it is it's it's amazing. It's an incredible experience that that no one else can match. Quest can a little bit, but but the the Vision Pro in part because [clears throat] Apple invented these cameras. Like Blackmagic is making them, but it's Apple's whole format. It's Apple like created the whole thing. And yeah, it like it's it just feels like this is a company that

it's like a lack of confidence.

Y

you let the device do what the device can do. Stop trying to like overdo it. put your we no step back. Yeah. Just step back and let it do its thing. And I'm more convinced than ever. I I again I've been having this take a few times. I wrote about the MLS thing. I wrote after the Metallica thing. I put on the front page of checkering today. I'm like look, no one in the world cares about this device other than me.

I know.

But [laughter]

I'm desperate for someone to read this. So there's no payw wall. You don't need a [laughter] forwarded to you. Just read read this. And and they had to do the Bucks, too. It's so funny. It's like the first one. It was a royal flush for you. I I have a meta question about just uh being in a position to review new technology and having to deal with uh like the Pepsi challenge. Are you familiar with the Pepsi Challenge story?

I am. I Dude, I'm Gen X. I'm very old.

Okay. Yeah. So, so the Pepsi challenge, for those who don't know, it they they they had people taste a little a 1 oz cup of Coca-Cola, 1 oz cup of Pepsi, and overwhelmingly the random people that came and tested said the Pepsi tasted better. And what the result was was that Pepsi was sweeter, so it tasted better over 1 oz. But over 16 ounces, a full can, it was more like 50/50. And people actually did prefer the Coca-Cola. And I feel like with a lot of hardware devices, you put it on for 10 minutes, it's amazing. You put it on for half an hour, it's incredible. But then then pretty quickly the you know the big tech executives say oh that's enough of the demo let's talk to you about something else uh now write your review based on 30 minutes and it's a very different experience than two weeks a month seeing if it collects dust seeing if you churned and I'm wondering just how you deal with that mentally where you have to talk about something but at the same time you're not getting the you know you haven't been able to spend a year with a product.

No it's a good point. So my way to deal with that is I generally don't do product reviews. So that makes it that makes it much easier. I think with the the vision pro though is a great example of this phenomena. It is like the first 30 minutes of using a vision pro is one of the most mind-blowing experience of your life and it continues to be. I actually I remember when I first got it I actually had uh I'm not I f perfect fall on from the watch segment cuz I found like a total douche this thing. I had my assistant I had it shipped to his house and had him fly to Taiwan to bring it to me because uh it wasn't it was only available in the US and the and and and

I remember actually I was going I think this is awful. This is so [laughter] patient. I was going on a ski trip like a week later. So I'm I'm up there and I remember I'm sitting in this bar up in Seikko and I'm like all the ski instructors like cuz I was friends with I had the same guy for many years. He brought all his friends over and they're all trying this. I like I hone my whole demo script and like we still talk about this afternoon with the vision pro in this bar in Deco trying it and it's it's amazing. It's absolutely incredible. But then

what do you do with it,

right? Like that's the big question. What I will say in the Vision Pros defense is even now I don't use it very often.

Every time I put it on though I'm like this is so cool. like it it retains that aspect of there's something magical about this experience. It's just searching for that ongoing reason to come back that use case and that comeback use case needs to be something that only it can deliver. The productivity things are cool. I I would I I but I I the other thing

the other thing I think is really real is like maybe from Apple's side like they have this device it has so much potential it's so exciting it's so cool and then it it it's almost disappointing that like okay the use case is just like live sports right because they're like we want to create the next computing paradigm but the reason that it actually is is an incredible niche is that like what's more mainstream than bread and circuses right what's more mainream

sports problem it's a chicken and egg problem you like the way you get those those productivity experiences and those new things that weren't possible before is you have a large install base that draws developers in to create those experiences. But to get there, it needs to be a market. So you need sort of the initial use case to create like this is why the iPhone is the greatest platform ever. Everyone needed a phone. Yeah. Like so like so you you got that built-in sort of advantage so that developers like they didn't have to do anything to earn developers. They just shipped the best.

Yeah. So that's what I'm saying like if you create this amazing live sports experience you get a million people that are using it multiple times a week and then and then they're using it. They're watching they're watching together. It's a social thing. Then they can be asking

social [laughter] I [clears throat] put it on. I don't look very social. No, but but I'm I'm talking about more like maybe you have a FaceTime call open with a buddy in the headset and you're watching it together. You're hanging out watching the game or something like that.

No, I I would just say I think it's the Vision Pro is more of a single player experience.

But what is the biggest single player market in the world?

Productivity. Like when I'm doing work like I want I I'm on my own working on it. Right now I have four monitors. Like I I'm one of those crazy people like a Vision Pro. I wish I would have brought with me last summer when I back to Taiwan because I was actually doing work and I was working on my laptop and it was terrible. Like, oh, I can't believe I didn't bring this. It was the perfect use case for me to use it. But by and large, I don't need it. But if there were particular apps or use cases that were uniquely enabled by it, maybe that would be a reason. But you're only going to get those by drawing in developers. You're only going to get developers by having an addressable market that is worth putting the investment into. Which means this is always been a this is one of my big thesis on strategy. It's always been a question on tech chicken or egg. What comes first? The device, the platform or the developers? I was at Microsoft in with Windows 8. I spent a lot of time and spent a lot of Microsoft's money getting developers. That doesn't work. What matters is demand. Demand pulls in developers. You need the core to sell a bunch of devices and then you can start the virtuous cycle. To me, the core use case for the vision pro anyone remembers it exists and I'm not sure that Apple executives do is live in your living room. You do have a bunch of devices and then you pull people in.

I have a buddy TJ who worked at Apple at one point when the vision pro was announced. He got it. He was so excited. He was like, "I'm gonna spend the next three years building products for the Vision Pro." How long did he last? He lasted like three months. Yeah. Like, no, no, no. Like less. He was like, "What's the point of build? What's What's the point of building for uh He's super talented, but what's the point of building if there's nobody there?" You know, you know, it's not

Are you Are you uh You mentioned the four monitors. Are you going in on the 52 in Dell 6K monitor?

No. Way too low resolution.

Way too low resolution.

Actually, I lied. No, I'm actually on the fifth monitor right now. Um, [laughter] just for podcast. So, I have I have two 4K monitors. I have my laptop. I have another I have with that LG like square screen. Super useful. It is Rosville resolution. But then for podcast, I have a 55 in TV here. So,

okay. Staying staying on Apple. Uh, Gemini the the Gemini Siri news dropped this morning. It had been kind of previously reported. So, in some ways it's old news. What are you What are you expecting out of like this new version of Siri?

Well, the bar is 55 ft underground, so [laughter] I expect it to be a lot better. Uh, no, I think it makes a lot of sense. I mean, it's uh [laughter] now that the federal courts have approved Google and Apple combining to rule the world, Google has Google has the infrastructure to support the scale of Apple. They know how to work together. They've worked together for years. They're are very natural partners in that respect that Google can think big picture about this. Like I'm sure Apple the report that Mark German had last fall was Apple's going to pay like a billion dollars which again we're in tech that's no money at all which but it's tied into the search deal like they can massage it. The problem with working with an OpenAI or or an anthropic is they need to make money and so like Apple doesn't I think there there's a more sort of I scratch your back your scratch my back sort of thing here. Google's talking about, you know, working with Apple's chips, adapting it, whatever it might be. Apple's does bespoke stuff like that. So, I I I think it makes a ton of sense for both sides. Um,

do you think do you think it ever flips? And do you think uh do you think Google will be paying Apple? Because there's this news also that you're going to be able to do shopping through Gemini. And so, you could imagine a world where you go to Siri, you ask for a product, and there's ads in that Gemini result. And and Google's the one that's monetizing that. So, they're passing some of that revenue back to Apple.

It's a good question. I mean, I think when Apple talked about like the next generation Siri and Apple intelligence, I was pretty optimistic about this idea of Apple basically replaying the search.

That's my that's my whole thing. I'm like I'm like in some way if I'm Apple and I'm paying for an LLM to use it to power a product that can basically do search really well, that could eventually have an ads business attached to it, that eventually could have like commerce built into it. Like it's this weird it's this weird situation where like Gemini is effectively helping like that's that's what I don't fully understand yet because I'm gonna be like

I think that the market structure was so perfect for Apple in that they need like it's just a default search engine. It's there's no sort of deep integration into the product. It's just like when you type in your search bar what engine is used and so it's super easy like there there's like a concept when it comes to like figure out who's going to win in a value chain. super easy subst sub substitility is good for whoever can plug whatever in. So they could choose whoever they want. So it made sense that the value flowed that direction. The difference with this AI stuff is it needs to be integrated deeply and Apple is more on the defensive here like they need to have quality AI features built into the operating system. And so I think the need is more on Apple's part. And the benefit Google is getting, my suspicion, is less that they're is less about the ad thing and more they're getting some incremental revenue and they're probably getting a lot more. I mean, Apple's going to be precious about the data. We'll see how that sort of works, but um but they also don't need it's all incremental.

Even like the internal reasoning rollouts and and all of that that happens in Gemini, like Apple's not going to be able to claw all that back. So you're going to have all this like reinforcement learning training on okay maybe maybe you're obiscating the the the the privacy data what the person asked for but all those interim steps of I went to a website I interacted with it I figured out how to click this button like that's that's going to acrue to Gemini you would imagine

well and what it speaks to I think is what would be the red flags that would come up to this deal from an Apple perspective it's like well they're not yet maybe they need to tough it out and build their own crappy LM so that in the future they control their own LM M such that if if you get to a world with say AR, right, where I'm I'm actually

lots of dispute about this, but I'm pretty optimistic on just in time UIs where the idea that something pops up that is nothing but the decision you need to make in that moment for whatever it might be and then it goes away. Like to me, this was the the best part of the Orion experience was the the Facebook sort of AR glasses was when I received a call because the the the OS I used there was kind of like Quest just sort of dumped in there and it would it didn't really make s you have like blocky stuff and there's like an Instagram and stuff. It's like I I could just look at my phone. This is going to be better here. But you could be

we got I got a conversation right now. I could have a notification pop right now saying so and so is calling and I could use my own little wrist thingy and and dismiss it and it's just there when I need it. It only has one option accept or decline and then it's gone and and I could see that being a future of UI and important for Apple to control but by not shipping their own by depending on Google they might say pretty words about oh we're going to simultaneously develop our own no that whatever if you're that's not going to happen. you're going to be so invested in this other one. And that speaks to the value is acrewing to Google cuz they're the ones actually developing and pushing the technology and Apple is sort of trailing along. Given that I have a hard time seeing you Apple's going to get more and more deeper into this using Gemini, are they going to be able to credibly go to Google say pay up or else we're going to switch to someone else? I think that's probably unlikely. So, I don't think it's going to play out like search, but we'll see. I was obviously wrong about this once, so I could certainly be wrong again.

Sure.

What was your reaction to the Manis acquisition? From my point of view is exciting. Um, specifically [clears throat] because Zuck has just been spending all this money on talent, but if you look historically, he's been very good at like buying something, scaling it. It's sort of unclear to me so far whether he plans to take Manis from a hundred million dollar run rate to multiple billions or just leverage the team's ability to build great agentic, you know, effectively workflows product. Um, how are you thinking about it?

Yeah, I I just I I have a hard time seeing the meta in the enterprise sort of angle. So

no but but I but even but even more talent than

but even it's like like figure out the best product in this category and buy it for me

like that that kind of work like that's what I can imagine the man being that's my sort of understanding yeah is like this is actually a really excellent product team that is doing very good work and is worth having on board and does that mean growing their business into something larger that's a material source to our business? I think that would be a mistake. Um, but I think the idea of agentic workflows uh is obviously a very compelling one. Actually, one of my favorite things Mark Zuckerberg has been all over the place on AI. I think I did an interview with him like nine months ago that was kind of bizarre. Uh, it was right when he was clearly thinking through maybe I need to reset everything and I thought that sort of came across in the interview at the time. But one thing he did say is actually what is the largest sort of most profitable atscale agent right now? Facebook advertising, which I think is a very or or Google advertising thing. A very astute point. You go in, you tell Facebook, I want

I'm willing to pay a $149 for a customer or $9.99 for a customer. Go get them. And it goes out and gets them. Now, is that a full LM denominated probabistic workflow? No, it's it's not really LLM's. It is more probabistically [clears throat] post. But the idea that you ask for a job, the the focus is on the job to be done as opposed to specifying how you do it and that is just by and large for most advertisers particularly in the long tail probably at part of the tail too they just don't want to give up control the better way to do it. And I think there's going to be lots of things like this. I think Google's announcement with Shopify and and this idea of of you know ads that are perfectly targeted the user very compelling makes a ton of sense. Uh that's where something like rock by the way fits in like super fast inference so you can generate like you think about how advertising works these auctions that are run and you get ads in the time it takes to load a web page is absolutely incredible where the the next step is would be insert generative ads into there which is going to require very high inference speeds. Um so I think and this has always been the most compelling short-term AI monetization opportunity is basically Google and Facebook ads. That's why I've been super bullish on both of them. I think it's why Facebook needed to do this reset um and why Google, you know, it's justifiably been sort of going to the moon recently.

Well, uh what do you think of Eric Seuford's point that uh that you might not actually need to put the ads in the LLM responses in the sense that you could be you could be going to an LLM and saying, "Tell me the history of the Roman Empire." It knows that you're shopping for shoes, for new shoes, and it shows you ads for shoes in the middle of your Roman Empire deep dive. Just like on Instagram, you can be scrolling one thing and your algorithm can be recommending you a certain type of content, maybe like dogs, showing you dog videos, but then it also knows that you need luggage for your next trip and it's showing you ads for luggage or something.

No, I think I think I think it's a great point. I think Eric makes some good points about like you're running a real risk of conflict of interest, even the perception of conflict of interest.

Totally. [clears throat] Yeah.

Otherwise, and so I think that's a very good point. And the reality is that's a much larger business like just being uh show what you're looking for is inventory. The number one way to predict

sort of the upside for Meta over the last several years,

the stock market has consistently had it totally backwards.

Every time they're sort of uh their price per ad would plummet because the stock market would freak out. This happened with stories in like 2018, 2019. It happened with with reals a couple years ago.

And what the what the the issue is that the reason why price brand plummets is because there's a massive increase in inventory. And when there's a massive increase in inventory, it's just more places to show ads. That is a huge opportunity. And you saw huge runups both times as they figured out how to monetize stories, as they figured out how to monetize reels. And this should be an opportunity of how to like it's it's going to take a while to figure out how to monetize you know maybe it's maybe just an aspect of when they're reasoning when they're thinking or image generation that might be an ad opportunity and it's a big opportunity a big problem is these folks I think the openai has hired so many meta people it's confuses me why they haven't been on board with this like it's a winwinwin you get you need money to fund your operation. The best way to make money is to is to show ads. Why? Because you get to deliver a better product to more people. This idea that we're going to commit to a world where if poor people get a worse product, that's not how tech works. And the reason why tech is amazing is it generates a ton of consumer surplus. You do this upfront investment and because it's monetized by ads, everyone gets the best product. It's great. and OpenAI like having this religious devotion for so long about not doing this. If they had launched, they could have launched the world's crappiest ads in 2023. By today, in 2026, they'd be good. They'd be making money and people would rebel against it. Now, they're going to have to launch ads. They're going to suck and people be like, "This sucks. I'll just go to Gemini or whatever it might be." just it drives me bonkers because like

it seems like there's a little bit of like uh not to go back to chicken and egg problem but like the fuh first mover disadvantage like the first LLM that has ads there'll be a whole press cycle about oh Gemini's the ad one

would have been a lot easier if you were the only LLM like that like the the it's it it's it risks the entire company like like they need to get their opportunities in the consumer space first and foremost because the consumer space needs to monetize via ads And the fact they didn't get there or start to get there, still haven't started to get there is uh it it's a company imperiling

poor decision.

Last question uh from my side at least uh predictions. Do do you think there's anything to the the prediction that OpenAI will buy Pinterest or partner up with a uh social network at a more deeper level than they already have?

Well, so it's really interesting. This is one of my uh I was actually uh uh I was talking to uh I've been met when I was in New York a few months ago and I was at this this off actually me and the semi- analysis guys uh great guys but they were sharing an office with like a hedge fund and one of the guys is like look you're responsible for one of our worst all-time decisions [laughter] and I'm like what's that he's like um we bought Twitter I'm like I never said to buy Twitter that's [laughter] terrible stock um unless he was going to overpay for it and he's like but then I remembered what it was it was I 2017 2018 when Twitter bought mopub.

Oh

and my theory at the time was Twitter's just a very poor inventory for advertising. Part of it is it's text based there at especially then there were fewer images.

There's also a mindset when you're on Twitter you're like girded up for battle and like you're trying to engage like opposed Instagram like Instagram the ads might as well be

the content

Instagram [laughter] like X is fight or flight.

Yeah. No. Exactly. Exactly. And so my theory then, but Twitter has the potential to really understand your interest in a really sort of deep way. And so what they could do is they could harvest signal on Twitter and they could manifest it with using Mopub and inventory sort of across a bunch of apps. And that would be sort of a very compelling model.

Again, who knows if I was right. Twitter's executives or Twitter was incompetent for years and years and years. At one point, I said, "I'm never covering this company again because this is pointless." Um, that's that's a bit what sort of Apploven did who did buy by by Mopub from from Twitter and has made a ton of money doing that. But I think that would be the thesis there, which is because we're people are dumping everything into this LM, we can get all this signal and understanding, what we need is inventory to monetize that signal and could that be inventory to do so? Um, the theory makes sense. It feels like they have a lot more important things to spend money on. I think probably at the end of the day, particularly post AT like O and O like your own and operated properties are always going to be the most valuable. I think figuring out how to monetize and jetp it's not an insane idea for that reason.

I like that. Well, thank you so much for hopping on the show on short notice. We're huge fans here and congratulations.

Look, I am on a look. No one cares about the Vision Pro but me. Therefore, I take it upon myself everywhere.

Correct the record and say the strongest soldier.

Well, camera.

Yeah. If you want the uh the most uh the deepest analysis, the most uh you know, the hottest takes on the vision pro, of course, sign up for strateg. [laughter] Let's do it.

Let's do it.

We need to We need to [laughter] Well, thank you so much. Have a great rest of your week and I hope your 2026 is off to a great start.

Yeah, great to see you, Ben.

We'll talk to you soon. Have a good one.

Label Box, delivering you the highest quality data for Frontier AI.