I also didn't see the notice it was a short story. So I was a bit puzzled when reading how Aaron built his own exchange as a novice programmer. To me building an exchange is
high up on the scale of programming difficulty. Its one of those things you have to do extremely well or else no one will trust your platform.
I found that some of the timeline seemed out for my worldview. I'm not really sure how long the whole AI bubble has been out there, but it just seems so recent that I wasn't sure how they fitted all that in in the current timeline. Perhaps that will mature, in 5 years it won't all seem so new and perhaps more plausible.
To address the GPs point though, I do think that with LLM-assistance even now you can certainly create such sites pretty quickly with a low skill level, and that will only become easier. The big problem with creating things (in that way, or manually) is that you are the architect and artistic director, and in truth many people just don't have the ability to manage the end-result in a way that will work. We see this in websites created by developers for their own software which get posted here. Too many times there are massive assumptions that the viewer knows all about the software, which means it is almost incomprehensible to the incomer who may be interested but leaves before engaging (that is: me). I don't suppose that was a problem for Aarons bot society, but it could have affected the launch.
I also didn't see the notice it was a short story. So I was a bit puzzled when reading how Aaron built his own exchange as a novice programmer. To me building an exchange is high up on the scale of programming difficulty. Its one of those things you have to do extremely well or else no one will trust your platform.