I have never used LLM's to write essays, so I can't comment on that.

What I can comment on is how valuable and energizing it is for me to cooperatively code with LLM's using agents.

I find it sad to hear when someone finds this experience disappointing, and I wonder what could go wrong to make it so.

I don’t thik someone finds this experience dissapointing but harmful for cognition, probably in the long run as the cognition ‘muscle’ athrophies in some regions as I see it. Remains to be seen how it pans out. However, how much would you be willing to pay for LLMs before you decide it’s not worth it? It is unexpensive at this stage but this won’t last.

	▲	grim_io 5 days ago \| parent [-]
		Unless they get Thanos snapped out of existence, I would probably just switch to inferior local models. Going back to pre-LLM is not an option for me. Not because I can't, but because I don't want to. How are you people using AI? I still have to think a lot. The biggest change is that I don't run around in circles trying to fix annoying bugs.