3 ms·
I have never used LLM generated code so far, but I would be extremely sceptical about letting an LLM generate code and using it without full review. IMO LLM gen
by voidUpdate 2y ago
I have never used LLM generated code so far, but I would be extremely sceptical about letting an LLM generate code and using it without full review. IMO LLM generated stuff should only be used as a starting point, not a final product
- shekkizh 2y agoAgreed. What would make you less skeptical? We typically do not look into the code within the software that we install in our systems today - so there is some level of trust already.
- voidUpdate 2y agoIf LLMs didn't hallucinate as much I'd maybe trust their output more. But they have a tendency to do unpredictable things, which isn't really what you want in a production environment. I keep seeing stories of what AI assistants have done incorrectly, which makes me hesitant to pipe their output into a codebase, unmodified
- JTyQZSnP3cQGa8B 2y ago> We typically do not look into the code I actually do it, and when I don’t there is a chain of responsibility to handle that. LLMs write random code that haven’t been approved, agents would make that worse by actually running it. Most AI fans want us to run commands like “cat /etc/random | sh” and that’s why software engineering is not real engineering.
- shekkizh 2y agohaha. I get that there are no rigorous standards in software engineering, but this has not stopped us from using software. I wonder if implementing a "chain of responsibility" -- kinda falls into my third scenario -- will improve trust levels for code generated by LLM.