That’s ok, but AI is useful in particular use cases for many people. I use it a lot and I prefer the Codex 5.2 extra high reasoning model. The AI slop and dumb shit on IG/YT is like the LCD of humans though. They’ve always been there and always will be there to be annoying af. Before AI slop we had brain rot made by humans.

I think over time it (LLM based) will become like an augmenter, not something like what they’re selling as some doomsday thing. It can help people be more efficient at their jobs by quickly learning something new or helping do some tasks.

I find it makes me a lot more productive because I can have it follow my architecture and other docs to pump out changes across 10 files that I can then review. In the old way, it would have taken me quite a while longer to just draft those 10 files (I work on a fairly complex system), and I had some crazy code gen scripts and shit I’d built over the years. So I’d say it gives me about 50% more efficiency which I think is good.

Of course, everyone’s mileage may vary. Kinda reminds me of when everyone was shitting on GUIs, or scripting languages or opinionated frameworks. Except over time those things made productivity increase and led to a lot more solutions. We can nitpick but I think the broader positive implication remains.

some people are so determined to be positive about AI that at some point it just comes across like they’re getting paid to be

It's very hard to see downsides on something like GUIS, scripting languages or opinionated frameworks compared to a broad, easily weaponized tool like generative AI.