>I get it, we can use AI agents, and they keep improving. Great, now calm down and go build stuff, and show me the stuff.

I completely agree with this sentiment. What I have seen is that these types of people never show the results of their work, they only discuss how productive they were or how fast it was to develop their project. The actual details of the project are not really mentioned.

In the rare event they do show their project, there's three categories: a worse version of something else that already exists (e.g. their version of Counter Strike that is less fun, rewriting some Python library into Rust but is now also worse), a dashboard, a visually engaging website you'll find interesting for less than 10 minutes and then never touch again.

I am not even an LLM hater, I am just not that impressed with majority of what has been built. Honestly I think back in the 90s I was far more impressed with the early websites that were totally garbage by today's standards. I remember spending so much time on the US Treasury's website as a kid learning everything about how currency works, I do not think I have ever had that level of engagement or fascination with anything that was generated by an LLM or where an LLM had significant support in the development.

Certainly this technology is nascent, however allegedly it should be making us far more productive so in the fourish years it's been around I feel like I should have witnessed something interesting by now. Perhaps I am a curmudgeon, but even in that case I think that there should at least be something else that's taking the world by storm and I am just not seeing it either.

Maybe there are some cool LLM creations, I think the speed at which it creates is also its own weakness. If anyone can create things really quickly, that includes people who are lazy and lack imagination, so then we're inundated with so much slop. Increasing accessibility does not mean that the quality will be maintained. Much like National Parks, if we make sensitive wilderness areas as accessible as possible that will degrade the quality significantly with the arrival of so many tourists, thus the parks have to consider the balance between preservation and conservation. Therefore, it was better when art et al. required some time, intention and skill before it could be shared with the world. These restrictions inherently kept the balance of quality much like how limits of accessibility in National Parks preserve the natural beauty.