Language models are exceedingly good at understanding the meaning of your language without the use of specific keywords. Here's an example from a recent search I did.
> metals can be flexed or bend, and it will regain some of its prior shape
in Google returns either "ductile" or "shape-memory alloy," which are both incorrect.
> What is the property of a material where it prefers to stay in its current form? This is often found in metals, where it can be flexed or bent, and it will regain some of its prior shape?
in ChatGPT correctly explains that
> The property you are referring to is called elasticity. Elasticity is the ability of a material to return to its original shape after being deformed (stretched, compressed, bent, etc.) when the external forces are removed.
We all know that LLMs can hallucinate, and they are therefore not a reliable source of truth or knowledge. I'm not necessarily trying to say that LLMs are more accurate than something like Google's knowledge engine. The value in LLMs is that they can infer your meaning to some degree of accuracy (just like asking a human) so that you can productively continue your own research in more depth.
I've found the same for setting up my environments. Before little things that were minor quality of life issues that I just didn't know what to search for, or couldn't be bothered to go hunting down in the docs, I can easily give an example and ask questions about. Little things that I put up with for years are all now resolved and I'm realizing how much nicer having a lot of little things not worth dealing with individually gone is. I don't have to put up with little things because I can't context switch for a day to figure them out, but I can spend 10 minutes with AI to resolve them without much brain power/context switching on my part.
In the IDE/tools/frameworks I use so many little things that 'it would be nice if it did this, but I don't have the time to figure out' are now resolved.