Clojurians Log v2

> The general consensus seems to be that the knowledge that these communities work in is hard-fought and the use of LLMs is a form of missing the point entirely. In these communities (keep in mind that there is an implicit “not all…” throughout) the process of mastering a difficult field itself is the product, and something that runs is generally a nice-to-have. This is a bulls-eye strike at how I feel, particularly with Clojure dialect development. Two years ago, anyone making a Clojure dialect was grinding through the work of understanding the intricacies of Clojure's JIT/AOT duality, appreciating the nuance of the various Clojure interfaces and the confluence of behavior they provide, and figuring out whatever host interop means for their platform/lang. Now, someone can port Clojure to a new platform/lang in a weekend and it'll pass the clojure-test-suite shortly thereafter.

I think these kind of project won't be treated as serious projects. It is really important understanding domain before advising using a LLM or not. If you want a mockup html/js/css it is really good to use LLM, if you want to write compiler or a software for an airplane control it is very dangerous to use LLM

Learning is actually a very valid usecase for LLMs, helping to shape mental models – but you still need to put in the effort, and learning-by-doing doesn’t work (it only works, as Fogus says, when you painstakingly follow the process, not when you teleport to the target). And of course you need to be wary of hallucination and validate the output. Here are two instances of where I found them helpful: learning https://chatgpt.com/share/6a731c01-5604-83eb-a5b8-cd1295d0efc5 and about https://chatgpt.com/share/6a743bc7-d0dc-83eb-acc9-8f2faaffc420