C (4:06)

Sure. I mean, I think that was a pretty good description. So it's, you know, imagine.

A (4:10)

Unbelievable.

C (4:11)

Unbelievable. You know, imagine like your phone, your laptop, but instead of everything being built around apps or icons, it's built around just a really capable AI that's running the show. So you'd interact with it conversationally, ask it to do things for you or connect tools for you, like you were saying. And those smaller tools would still exist on their own, but the apps would plug directly into the AI. So it'd be like, hey, I need to plan a trip. And the system can just pull up a calendar, pull up a flight map automatically, and the device itself would work more like a living assistant than a piece of hardware with apps just slapped on top of it.

A (4:45)

So WeChat, the Chinese super app, has mini programs, and that's kind of been the gold standard for a lot of folks in the west trying to create a super app. Lauren Good and Will Knight of Wired say if the web and mobile eras of the past 30 years were defined by users browsing the web or being locked into a mobile app experience, OpenAI is now combining the two into its own kind of chat driven OS. Grace, you, you wrote OpenAI introduced this wide swath of app integrations for ChatGPT, pushing gen chatbot towards super app status. Do you think this is the super app that west has been waiting for?

C (5:25)

I think that it's on its way there. I don't know that you could at all compare it to WeChat at this point, but I think that OpenAI is definitely pushing in that direction.

B (5:32)

Okay. And I think, I mean, It's a great analogy with WeChat, but I think the big difference is to WeChat was established during an era where there wasn't a really vibrant Internet in China, was people had mostly were connecting through phones. This was kind of, this really almost de facto was the operating system for phones in China and took on all these mini apps. OpenAI is coming into a, into a ecosystem which is a lot more developed. So the web is still going to continue, apps are still going to continue. I think OpenAI also wants kind of to connect to those and connect those to their super apps. So it'll be an operating system around the app, but also will connect to these other apps through APIs and other apps in the Internet through APIs.

A (6:20)

Speaking of these apps, Grace, which apps do you think will become the most useful or popular inside ChatGPT?

C (6:29)

I think there's going to be a really big focus on productivity and basically the ones that can act like mini experts. So scheduling assistance, trip planners, personal finance bots, things like that. Tools that can handle a specific task and then use AI for the heavy lifting. I think the bigger part is going to be when they're able to start connecting with real world context. So your camera, your voice, your location, things like that. Maybe it can remind you to leave early because of traffic on your way to work or summarize your inbox before you open it. That latter one is they're both on their way. There's a lot of summarization tools. But yeah, I think I just early on we'll see a blow up of productivity tools. Just anything that helps people work faster, stay organized.

B (7:08)

Yeah.

A (7:09)

Yuri, what's top of your list?

B (7:10)

I mean, I think, I think Grace is spot on there. I think also just vertical information providers. So Zillow I think is a great example. You know, they are, you know, they provide real estate information. They're really deep on what houses are for sale, where they are. That type of app is, I think, I mean, it is already an early app, but I think that category of apps are going to grow rapidly.

A (7:36)

So it sounds like they want this to replace Google as kind of the gateway to the Internet. Head of Chat cbt Nick Turley did say though that he said, will people spend all of their time in Chat cbt? I don't think so. I can imagine you starting your day with ChatGPT, then being guided toward other apps and websites, going on to say not all interactions with the commercial world need to be a chatbot. Where does, where does the traffic come from? Is it Just people not using Google Search as much and starting their Internet experience in ChatGPT. Is it just going to be kind of a migration that's that linear?

C (8:19)

I think we're on our way to a really slow migration because people are not just not using Google Search. Overall, people are not walking away from Google Search. They're not walking away from traditional search engines. While there's a pickup with AI search, it is not anywhere near replacing it. There's a huge customer base, there's a huge user base, there's a lot of demand for these tools. But Google Search is not out.

B (8:39)

Grace is absolutely right about the current state. And even in the future, the long tail. So referrals are going to go down from search to a lot of sites. The long tail of the web is going to be hit hard by that. But the sites that have dedicated and loyal users, you know, if you're going to New York Times directly or Zillow directly, people will still go to that directly if they're, if they have a connection to it. So it's, it's not going to. People will still go directly to these sites even if they've migrated a lot of their behavior to Gemini or ChatGPT or whatever other gen AI program.

C (9:19)

And you know, there's AI itself within Google search, AI overviews so that, you know, for better or for worse for referral traffic, but there's AI users who are simultaneously Google Search users.

A (9:29)

Yeah, one of the interesting questions here for me was the kind of data sharing piece here Max Zef, sorry, Maxwell Zeff of TechCrunch again was saying, was talking about how the data sharing piece will work. Writing key questions around apps in ChatGPT will be privacy and how much data third party developers will have access to. OpenAI says developers must collect only the minimum data they need and be transparent about permissions. However, it's unclear whether developers would have access to a user's entire conversation with ChatGPT the past few messages or just the prompt that summons up the app. So I think that's a really interesting piece of this because that could be a significant amount of data that you're getting from someone if you're getting whole conversations versus just a couple of messages.

C (10:15)

Meta is on its way to doing that. Using conversations with AI bots to train its data.

A (10:20)

Okay, so maybe. So you think they'll probably more than likely be heading in that same direction.

C (10:24)

I think OpenAI, I think that they are perhaps moving a little bit more slowly with their public policies around user privacy data just to think around out of concern and they don't have as ingrained of a position in people's lives as meta's platforms do. So I think that it would be easier to scare people off. I think people with meta platforms have already kind of accepted that a lot of their data is going to be used for ad targeting, for personalization. I think that it would be easier to scare people off with straight up AI tools like OpenAI if they implement those policies too quickly.

A (11:01)

Yeah, big question here for me is the revenue model. How are they, how are they going to make money? I was reading Economist article saying that they're expected to lose money off of $10 billion worth of revenue this year. And so a lot of folks, particularly shareholders, wondering how are you going to make money? And there wasn't any details around the revenue sharing agreements with these partner apps, but they just introduced instant checkout. So maybe buying things through the chatbot, through, through these apps and also just surfacing relevant apps when the user asks for something. If you ask for a vacation, homestays open. Sorry. Airbnb vs. Vrbo could bid for priority ranking. A lot of folks using those types of. There was this chart on the screen you can see from Critio showing that close to 40% of Americans already say they use AI for accommodation suggestions. 20% use AI to plan their entire trip. So you could easily see a world where you're in ChatGPT and you're asking these questions about stays and these different companies are competing to be the first one that pops up and helps you to plan that trip. Any thoughts on how the revenue sharing model here?

B (12:19)

Yeah, So I think OpenAI and, and probably other companies are going to have a mix of monetization strategies. The revenue sharing piece is going to be part of it. I think they're moving that that way with the, the, the buy the, the commerce integrations that they have. I think they're going to explore affiliate model a little bit more than they have. They'll probably almost definitely use advertising. I, as I'm, I'm doubtful that they're going to do preferred integration into the actual answers because so much is based on the trust of the answer as an agent. So I think they're going to be very cautious around doing that. Google may go a little more quickly into that but, But I think OpenAI is going to be cautious around boosting recognition within an answer, but I think their advertising is going to be around those answers and I think there's no doubt to, to cover the costs that they have to cover, they're going to have to do advertising.

A (13:24)

We touched on Google briefly, saying Gracie was saying that there's still a lot of people who are, who are using Google and it's not this kind of great migration that we're seeing. Jeremy Khan, Fortune's AI editor, was saying that this is the great platform shift people have been anticipating since ChatGPT debuted. Now the question is how quickly and how completely consumers and businesses will move to this model. Out of 10, how likely is this to disrupt Google, let's say this year and then how much? So in 2026, I think 10 being.

C (14:03)

Total disruption for this year, I'd say for OpenAI and Google, maybe like two and a half. I think it's going to be a really, really gradual change. I think that a lot of the stats that you see about however many high percent of people are planning their trips or doing all their shopping with AI, that's usually out of gen AI users, not out of general consumers. So the people who have adopted it have adopted it really hard. But it's still only on a slow, a slow momentum. 2026, again, I don't think that in the next year we're going to see things change quite quickly as they did from 2022 to 2023 when all of these apps really broke onto the market. Like I said, I think it's going to be a pretty gradual change and it's going to take time for that trust build up. As we see more of these app integrations coming in, the platforms become a lot more useful and I think that that is going to boost these cases a lot of.

B (14:52)

Yeah, Yuriyg, I agree with 2026, I think it twos sounds about right to me. It's going to happen an impact, but not a very big impact. And I'm sorry, in 2025 I think it'll have an impact, but not a very big impact. I think 2026, I think it's going to be gradual until it gets. It'll hit some sort of inflection point. So I could see by the end of 2026 actually having a significant impact in the way we interact with the Internet and with commerce. That said, it'll disrupt Google's current business model, but I think they're also pretty well positioned to be one of the winners in the new model as well. So Google as a company may actually do very well, but the old CPC model might be significantly disrupted by the end of 2026.

A (15:38)

So Google has an AI mode, it has Gemini. Will Google copycat this Will people just do the same thing, you know, opening up apps within the Google AI model?

B (15:52)

I think so. I, I think Google will start with its own apps. It already has a little bit. But you know, I think much better maps integration, much better shopping integration I think is a no brainer for Google. I think they're going to move quickly on that and it makes complete sense they're going to follow with other apps as well.

A (16:12)

Brian X. Chen and Trip Mickle of the New York Times writing about the modern AI assistants, which are far more capable and flexible than the clunky voice helpers like Siri, saying they're poised to become the central OS of our personal computing devices. Focusing a bit more on the device piece here, superseding smartphone software in importance, experts say they write that apps and their polished interfaces or won't matter much when AI assistants use devices on our behalf, automatically carrying out tasks like making plans with friends, generating shopping lists and taking notes and meetings that would spare us the need to swipe through software menus and type on keyboards. They list AI devices in their article like smart glasses, ambient computers, reimagined smartwatches. And then Dan Gallagher of the Wall Street Journal was writing about the same topic, consumer AI Devices, saying that Meta apparently believes that Apple's dominance of consumer devices is vulnerable in the AI age. Boss Mark Zuckerberg sees smart glasses as crucial to his vision pun probably intended for an amped up version of AI called superintelligence. Yuri what might this OS competition mean for consumer AI devices?

B (17:29)

I think AI voice based devices are going to completely run on these AI operating system. So you know, you see it in Alexa already with Alexa plus. But I think cars, I think smart glasses are really going to work heavily with these AI interfaces, not just through voice but through hand motions and things like that nods phones as well. So Siri or some replacement of Siri is going to be AI operated as well. And I think it does really put Apple in a vulnerable spot. That said, I think phones are still going to be the center of our digital life for a foreseeable future.

A (18:14)

Grace I don't understand why so many people are trying to replace the smartphone. I know it makes sense if you're a company because you want to be the person that sells the smartphones, but it seems like it seems like no one is asking for this. The smartphone works perfectly well. It's a tiny screen, so having a pair of glasses on your face, maybe you can talk to it and it relays information. You can live more in the Real world, but you don't have that kind of visual aid that the tiny screen in your pocket offers. Is Apple vulnerable here at all in your opinion?

C (18:51)

Well, I guess to your first point, I don't agree with that. I think that there is a lot of demand for these, for these new form factors, for these new wearables. I mean if you're looking at how the Ray Ban meta smart glasses are selling, they're really selling. And there is I think a lot of demand for these new devices where it's central in conversation instead of catching and swiping. With Apple, I think vulnerable was a good word. It definitely puts pressure on Apple, especially with how slowly Siri developments have been moving along. You know, right now everything on your iPhone goes through iOS but people start spending more time talking to an AI that runs independently, then that balance shifts. If OpenAI or anyone else builds an AI first operating system, I think that kind of rewrites the Playbook and then it's not about iOS versus Android, it's about who has the most helpful AI layer. Apple's still in a really dominant position. I mean it's got great hardware, it's got loyal users, it's got a really tough to lead ecosystem. So for Apple I think a change in power would be really gradual. But I think over time people are going to start choosing their device not just based on camera or processor, but about who has the AI ecosystem and assistant that they prefer the most.

A (20:01)

Interesting.

C (20:01)

But yeah, I think there's a lot of demand for these new form factors. I think the one that's hard to pitch probably would be a screenless non wearable device. Just because you're not going to want to carry that around in addition to like some sort of pendant without a cord, something like that. It's going to be really hard to pitch people to carry something around that's similar to the function and size of their iPhone when you already have on device AI.

B (20:25)

Yeah. And I mean Sam Altman's mystery device is reportedly one without a screen. I agree with Grace, that's hard to see how that would take off.

C (20:35)

And he hates smart glasses. He already said it, so probably won't be that.

B (20:38)

Yeah, and Apple does have the dominant position now and their vision for AI, you know, Apple intelligence is, is really compelling and I think if they can deliver on that, they're going to be in a great spot. They just haven't shipped it, they haven't delivered on that vision and that's, I think what's leaving an opening for others.

A (20:56)

That'S unfortunately where we have to leave the conversation for today, but thank you so much to my guests for being part of it. Thank you first to Yuri.

B (21:03)

Great to be here as always.

A (21:05)

And of course to Grace.

C (21:06)

Thanks for having me guys.

A (21:07)

Yes indeed. And thank you to the whole editing crew crew and to everyone for listening in to behind the Numbers, any market video podcast made possible by Fetch. Make sure you subscribe and follow and leave a rating and review. We'll be back on Monday. Happiest of weekends.