>>105737418
The post I'm responding to seems like bait, but it isn't. Yes, prompt engineering has a bad reputation because pajeets love to throw it around along with the latest buzzword, however prompting is a skill. If you're incapable of explaining your ideas in a digestible manner (i.e. step by step) taking into account the limitations of this technology, you won't get anything out of LLMs. This will create a gap between you and people who use this with a background in programming. It's not surprising that some here can't even output a single method without having the SOTA models fuck up badly, while others can design a perfectly functional complex CI/CD pipeline and code within hours.
