I'm not sure I buy the "everyone will be AI coding to replace things that cost money with their own apps" idea. I only have so much limited time in my day (and only so many tokens on my claude account per week). It's probably going to make more sense for me to buy a tool that's been given human attention over the span of weeks over something i prompt into existence in a few hours (especially if I need 10 such tools to accomplish something).
"The economics of opportunity cost are unchanged" a friend told me recently, and I think that's exactly what is driving your intuition here as well.
I can’t always see the personal appeal, however when I view through the lens of businesses that buy very expensive enterprise software and other SaaS products (maybe blending into consumer market), well I think they’re toast. I think the acceleration of AI tools recently isn’t going to be indicative of how long the full transformation will take, but a lot of companies will start preferring Build over Buy. I have no idea the scope, but this is already happening at some partial scale.
I agree the free money in like one month coded SaaS apps are in big trouble. But like there's no way I'm gonna have the vision to desire to play a game I made myself using AI for instance (just the fact that I prompted it into existence ruins some of the exploration of a game made by someone else). So at the low end of the extreme (easy to make SaaS apps with basic code and a db) AI is a thread, but at the other end of the extreme (requiring vision and where human attention is a bottleneck) there's definitely still tons of opportunity.
I think a lot of people will do this, it remains to be seen how the actual economics of this shake out in the long run, especially considering, it’s not like the existing vendors are going to remain static.
The choice isn't between A) the expensive, proven tool and B) the thing I promoted in a few hours - it's b/w A, B and also C) the less expensive, somewhat proven tool that someone else prompted over a couple of days. I can see, over time, a slow drift towards "free".
Factor in how a lot of tools have weaponized their interfaces against their users - then the motivation isn't just cost, but usability.