3 ms·
it's just a consequence of the annoying anthropomorphizing of tech. "Human's can't do X, the AI model can't do X, look it's just like me fr, fr". Of course nobo
by Barrin92 3y ago
it's just a consequence of the annoying anthropomorphizing of tech. "Human's can't do X, the AI model can't do X, look it's just like me fr, fr". Of course nobody applies that logic to a forklift or a debugger. If gdb started to hallucinate variables into existence we don't call it a creative act, we call it a bug.
Given that these AI systems just like any other machine operate at scale, automated, and fast, they must be precise and transparent, that is where the work should be.
- pixl97 3y agoI don't really think you understand the premise of generalization. A forklift is not a generalized machine. It is a specific machine for a very well defined set of tasks. A LLM can take a set of input information, choose a number of options, like using an external tool, and act upon that data.
- nomel 3y agoI'm trying to understand this type of perspective, which is somewhat at odds with mine, so the following questions are completely genuine: Which models have you used? What did you try with them? Have you used GPT-4? How frequently?
- Barrin92 3y agoI use Copilot pretty frequently when I program, find it very useful to explore an API i don't know or just help write bash or other stuff I suck at or just to save me some typing. Still often produces very unidiomatic and non-performant code though. But it's worth it for what it costs. I do have a separate GPT-4 subscription as well but I've used it less and less. Tried to use it for search and text summary but it makes too many mistakes. Invented authors, papers, wrong summaries are just too common for it to be useful. So effectively I found myself Googling every time i used it to make sure the information is correct, which made the entire thing redundant.