Post #3004785

I am thinking to stop using LLM completely which I have tried many times previously but somehow I end up there again. But this time I am determined. Thinking of which, a few things come to mind. It's true that LLM hallucinates, and create their own data which in turn produce incorrect information. But, they are also quite helpful for some tasks like: recently today, I was updating my Hyprland setup, which produced a stuck screen with some errors. I uploaded it's image to Gemini and it quickly got the root cause. Further, they are useful say to quickly understand a long file of poorly structured code. So, I am kind of in a dillema. If I completely give up on LLMs, then I would effectively "waste" (is it?) my time tackling these sort of tasks. But, if I used AI for these tasks, I couldn't help but use it for tasks which I shouldn't be using LLMs for (like writing code, getting answer to some physics question I am stuck at for days, other creative tasks etc). Can someone help me out on this, please? #physics #AI #llm #cs #advice