This statement, "AI is often an accelerant more than a disruptor, and large companies benefit from existing customer relationships, strong brands, and lots of data." Does this assume that the fundamental customer niches and the nature of the problems being solved will not be significantly altered by the arrival of AI? The danger point for incumbents comes when new tech changes the framing of what the market / need states even are at a fundamental level. At that point, incumbency becomes dead weight. Are we sure this is NOT going to happen here? I honestly don't know; I've not thought about it enough yet and it's probabaly different by vertical.
Closest to the consumer tends to hold the most margin. ESP compared to manufacturing. (NVIDIA is a rare exception).
And specific use-case startups have reliably been able to create venture returns competing with the googles and facebooks. So I think there’s plenty of space in the apps side.
That said, I think installed user bases wins out and big tech gets bigger with a few new big tech winners like clause and got and grok.
👏 Finally, someone naming what most don’t want to admit.
The app layer’s running on borrowed time — orchestration hacks, webhook glue, and brittle integrations that won’t survive the move to autonomous execution.
The real investable bet isn’t another interface — it’s rebuilding the infrastructure layer AI will actually demand. Event-native, identity-resilient, and built to coordinate—not just compute.
I’d love to hear about companies who are building the way you suggested at the end of the post. Do there become ui specific solutions or does the user just ask ChatGPT to create them for them or do we just talk to a ChatGPT and it does everything that was once ui
Thank you for sharing your perspective. And I appreciate the clarification on the term App. Most disagreements are the result of misaligned definitions.
I'm curious regarding your perspective on the buy side of Large Enterprise vs. SMB.
I can see the argument as presented for Large Enterprise figuring out their changes to the stack, if anything.
As for SMB, I'm uncertain. Or maybe there's utility for SMB, but its not enough to worry about for investors? I'm curious.
I think viewing the app and infra layers as separate probably creates a false dichotomy. In today’s AI marketplace, companies almost always need to address both, and the line between them is blurred IMO.
There is no “AI app layer”. There's just apps. LLMs are just next generation databases, a way of storing and retrieving information.
This statement, "AI is often an accelerant more than a disruptor, and large companies benefit from existing customer relationships, strong brands, and lots of data." Does this assume that the fundamental customer niches and the nature of the problems being solved will not be significantly altered by the arrival of AI? The danger point for incumbents comes when new tech changes the framing of what the market / need states even are at a fundamental level. At that point, incumbency becomes dead weight. Are we sure this is NOT going to happen here? I honestly don't know; I've not thought about it enough yet and it's probabaly different by vertical.
Love 🥰 your ‘Headless App’ idea 💡
Closest to the consumer tends to hold the most margin. ESP compared to manufacturing. (NVIDIA is a rare exception).
And specific use-case startups have reliably been able to create venture returns competing with the googles and facebooks. So I think there’s plenty of space in the apps side.
That said, I think installed user bases wins out and big tech gets bigger with a few new big tech winners like clause and got and grok.
👏 Finally, someone naming what most don’t want to admit.
The app layer’s running on borrowed time — orchestration hacks, webhook glue, and brittle integrations that won’t survive the move to autonomous execution.
The real investable bet isn’t another interface — it’s rebuilding the infrastructure layer AI will actually demand. Event-native, identity-resilient, and built to coordinate—not just compute.
I agree. And though you mention how other VCs view this, I also want to point out the media narrative from most outlets is missing it too
I’d love to hear about companies who are building the way you suggested at the end of the post. Do there become ui specific solutions or does the user just ask ChatGPT to create them for them or do we just talk to a ChatGPT and it does everything that was once ui
Yes, I agree with you. I wrote about this a while back: https://open.substack.com/pub/davefriedman/p/sequoia-still-thinks-ai-is-software?r=37ez3&utm_medium=ios
Thank you for sharing your perspective. And I appreciate the clarification on the term App. Most disagreements are the result of misaligned definitions.
I'm curious regarding your perspective on the buy side of Large Enterprise vs. SMB.
I can see the argument as presented for Large Enterprise figuring out their changes to the stack, if anything.
As for SMB, I'm uncertain. Or maybe there's utility for SMB, but its not enough to worry about for investors? I'm curious.
I think viewing the app and infra layers as separate probably creates a false dichotomy. In today’s AI marketplace, companies almost always need to address both, and the line between them is blurred IMO.