it seems curious to me that you think vertical market software is at risk and yet you also believe the deeply embedded software that has to be deterministic and has higher switching costs will be the least disrupted, which i believe to be vms. am I reading it wrong? great work regardless of this question i had.
it seems curious to me that you think vertical market software is at risk and yet you also believe the deeply embedded software that has to be deterministic and has higher switching costs will be the least disrupted, which i believe to be vms. am I reading it wrong? great work regardless of this question i had.
Indeed. Check out this interesting chart from Iconiq State of AI (Jan 2026):
AI builders (~69%) are most focused on building vertical AI applications, up from ~59% two quarters ago.
This is now much higher relative to horizontal platforms, platforms/infrastructure, models, and consumer AI.
https://x.com/EugeneNg/status/2019315659101134949
thanks Eugene. that’s interesting. we will see how well they can compete.
SaaS with data repositories are very hard to replace,
Also API integration with networks can’t be vibed coded because it is proprietary.
The problem isn’t the use case, it’s a story of multiple compression.
Excluding mega caps or large caps in the saas space, most are still trading at 30x + multiples after drawdown.
Maybe we get closer to 20x, there’s enough margin of safety from an investors point of view to offset disruption, if any, by Ai.
General software if dead. Find the niche.