Maybe everyone should use the term LLM (large language model) instead of AI for all such projects. But I guess then nobody will get hyped over it, especially investors who wish to make quick money out of the current trends.
@nixCraft Say what you will about AI, but it's allowed the "Well ackshually" demographic to streamline their undeserved confidence by outsourcing their habit of getting basic details hilariously wrong.
@nixCraft I keep trying to tell people this. LLMs ARENOTAI! We must stop calling them that. LLMs will never replace that which is closest to AI. They fail in some very basic ways that prevent them from even being able to do the same things. The best they could do is act as a side mechanism for complex algorithms that were already closer to real AI than LLMs can ever be. (Not even a backend.)
I'm very worried about the way everyone is determined to think LLMs are AI.
@nixCraft I use the term „autocomplete tools“ and actually my management is even more happy as they finally understand what’s going on and how to use AI 😅
@nixCraft The truly maddening thing is that science is one of the few fields that LLMs actually have potential to be useful in. The problem, of course, is you can't use it as an AI because it is not an AI. You can, however, use it as a mechanism to do things such as classify and organize patterns (lots of potential in biology especially!) Everything has to be double checked and it can still miss stuff completely, but it could speed many processes enormously as a starting point.
@nixCraft I'm not sure this says anything. If I bake a cake that doesn't taste right I learn from it. Just as ai is in a early state now doesn't mean it won't fix these problems.
Add comment