Thanks Mert. That's mostly my point, though I don't think it's fair to say it was *only* a distribution rail, hence my three points on what mobile was. And then singling out the importance of how it dramatically expanded the amount of time ppl spent online. (Your distribution point.)
Great article, thank you! Very interesting and helpful framework to think about the AI opportunity versus mobile.
I’ve been thinking a lot about how these AI start-ups justify their valuations and create monetization going forward. There is the obvious AI engine, but in the near term I think their best bet is to innovate a new phone, earbuds, devices, Alexa, and implement their ai into it, and take market share from Apple/Amazon. It’s a huge market and ripe for disruption because Alexa and the iPhone have had pathetic levels of innovation for years. I wouldn’t be shocked if Facebook and open AI are in stealth mode for phones. Unpopular opinion I know. Open AI would be at an advantage because of brand recognition among young people. Jony.
Great writeup! The chart about time spent alone got me thinking:
The mobile wave primarily produced isolating applications that detracted from human-to-human connection. If you extrapolate this chart into the future, you’d expect to see the lines continue to trend upward.
But, could it be possible that AI will actually promote connection, and what would that look like? With the so wary about connection by being replaced by AI, I could imagine many AI-native “multiplayer” apps being built with it as a core value.
This thought exercise is partially me trying to see how we can continue spending more minutes with technology WITHOUT losing our minds 🤷 possibly be wishful thinking but it’s a future worth aiming for!
I agree with you on one level - today’s business models clearly reward solo engagement. But multiplayer AI apps will build network effects that could become powerful moats in an environment where others types of moats are becoming less relevant.
For example, an AI language learning app that leverages human-to-human social dynamics (e.g. Duolingo) will beat out a competitor that offers cheaper and/or more powerful solo practice. An AI networking platform matching humans based on interests will be superior to a single player AI business coach.
They’ll need to nail the single player experience to start, but for longterm retention it feels like they’ll need to build those network effects.
While yes it competes with social media, it can still steal that share of time and grow the pie of minutes from offline time. AI has been adopted much faster since these devices are already in place but its usage leads to less in person social time which is replaced with at home time talking to AI.
The character AI reddit is also pretty wild - people spend 10+ hrs, talk to therapists about how to reduce their time etc. There's already been self-harm lawsuits related to this:
I think we will look back and realize that while AI eliminated jobs the bigger threat to the next generation was how it completely upended their social lives. As a mom of two boys I worry a lot more about this than the job elimination threat.
meant to comment this but emailed ahah
Mobile, as a wave, was a distribution rail. Hence increased the availability of experiences leading to more time spent.
AI has no such effect (maybe Siri like AI) - because it is purely a capability
Just wrote about this today so could not resist asking ahah
Thanks Mert. That's mostly my point, though I don't think it's fair to say it was *only* a distribution rail, hence my three points on what mobile was. And then singling out the importance of how it dramatically expanded the amount of time ppl spent online. (Your distribution point.)
Where is your post?
Time spent is indeed a good way to look at it - I try to look at it as new vs old experience more but it might just be semantics.
You can find the post here: https://aimode.substack.com/p/we-forgot-how-to-build-with-ai
Really great write up. The utility of a technology always comes first, then the minutes. At least that was how it plays out every time.
Why would someone want to build an AI companionship startup? Why would you want to work with someone who wants to build an AI companionship startup?
Do chatbots and AI products strike you as even a fraction as entertaining and useful as mobile applications?
Great article, thank you! Very interesting and helpful framework to think about the AI opportunity versus mobile.
I’ve been thinking a lot about how these AI start-ups justify their valuations and create monetization going forward. There is the obvious AI engine, but in the near term I think their best bet is to innovate a new phone, earbuds, devices, Alexa, and implement their ai into it, and take market share from Apple/Amazon. It’s a huge market and ripe for disruption because Alexa and the iPhone have had pathetic levels of innovation for years. I wouldn’t be shocked if Facebook and open AI are in stealth mode for phones. Unpopular opinion I know. Open AI would be at an advantage because of brand recognition among young people. Jony.
Great writeup! The chart about time spent alone got me thinking:
The mobile wave primarily produced isolating applications that detracted from human-to-human connection. If you extrapolate this chart into the future, you’d expect to see the lines continue to trend upward.
But, could it be possible that AI will actually promote connection, and what would that look like? With the so wary about connection by being replaced by AI, I could imagine many AI-native “multiplayer” apps being built with it as a core value.
This thought exercise is partially me trying to see how we can continue spending more minutes with technology WITHOUT losing our minds 🤷 possibly be wishful thinking but it’s a future worth aiming for!
Agreed on it's a future worth aiming for! Def possible, though incentives don't feel aligned.
I agree with you on one level - today’s business models clearly reward solo engagement. But multiplayer AI apps will build network effects that could become powerful moats in an environment where others types of moats are becoming less relevant.
For example, an AI language learning app that leverages human-to-human social dynamics (e.g. Duolingo) will beat out a competitor that offers cheaper and/or more powerful solo practice. An AI networking platform matching humans based on interests will be superior to a single player AI business coach.
They’ll need to nail the single player experience to start, but for longterm retention it feels like they’ll need to build those network effects.
I agree AI for consumer use will accelerate when hardware comes but I don't think the adoption shift being as big as mobile is gated on this.
The dopamine and addiction risk for parasocial relationship with AI and AI characters is already being recognized in studies and usage of these tools.
Here's a couple of the studies that look at this: https://www.sciencedirect.com/science/article/pii/S0001691824005006?via%3Dihub
https://link.springer.com/article/10.1007/s44230-025-00090-w
While yes it competes with social media, it can still steal that share of time and grow the pie of minutes from offline time. AI has been adopted much faster since these devices are already in place but its usage leads to less in person social time which is replaced with at home time talking to AI.
The character AI reddit is also pretty wild - people spend 10+ hrs, talk to therapists about how to reduce their time etc. There's already been self-harm lawsuits related to this:
https://www.reddit.com/r/CharacterAI/comments/1abub22/characterai_users_a_question_for_you_why_are_you/
https://med.stanford.edu/news/insights/2025/08/ai-chatbots-kids-teens-artificial-intelligence.html
I think we will look back and realize that while AI eliminated jobs the bigger threat to the next generation was how it completely upended their social lives. As a mom of two boys I worry a lot more about this than the job elimination threat.