Comment by lvl155
1 day ago
Sorry to be harsh but that just sounds ignorant. LLM can be a great tool to speed up your learning process especially for devs. It can help me get over blocks that used to take me half a day of digging through docs and codes.
Having a quick answer is different than internalizing some domain, especially when you may need to argue about it or propose some novel method. The former can work if you know the right question, but that's just as difficult as knowing the answer.
And most things that are useful daily is not pure knowledge. It's adapting the latter to the current circumstances (aka making tradeoffs). Coding is where pure knowledge shines and it's the easiest part. Before that comes designing a solution and fitting it to the current architecture and that's where judgement and domain knowledge are important.
I may sound ignorant but if you delegate your knowledge to LLMs you'll stay ignorant. Learning is a never ending journey, you can chat about a problem with a non deterministic stocastic compressor with stale knowledge, no problem but don't call it education. Core domain knowledge still makes people successful not chatting skills. Sure if you have deep domain knowledge you can still benefinit AI filling the tedious details, I'm not anti AI
> I may sound ignorant but if you delegate your knowledge to LLMs you'll stay ignorant.
That's kind of an ironic statement given the context. AI is just a glorified search engine that makes it very easy to find relevant information on a topic, just like a search engine but faster. One must still verify the results to be true, just like a search engine. AI is a tool to help you do your work faster, not do it for you, and should be trusted as much as any other anonymous source.
Search engines index content, LLMs "compress" content during training and regurgitate it back wrong or in non deterministic way