@gmerc@DBKJanus@gardenGnosticos you can have different opinions but stating them as a fact + attacking other people for it makes your opinion hard to respect
@LostKitsune31@drgoononyou@Burgerman42069@14ChikaWham there are fewer issues with AI code than human code, especially regarding deterministic outputs. a good SWE managing AI agents properly will always be better than if they did it without AI. they just produce raw code quicker. also what’s your CS background like?
@LostKitsune31@drgoononyou@Burgerman42069@14ChikaWham I’m not denying your anecdote but i’m telling you as a SWE that LLMs are just better coders than majority of us. manual coding is effectively obsolete and 99% of tech companies will expect you to use AI.
@LostKitsune31@drgoononyou@Burgerman42069@14ChikaWham every project made purely by human skill does not work perfectly. if that were the case then bugs, exploits, debuggers wouldn’t exist. also AI is primarily used to decomp, where they can literally check the output bit by bit to ensure accuracy
@LostKitsune31@drgoononyou@Burgerman42069@14ChikaWham this is just not true anymore man. i thought the same before but nowadays, LLMs are just better coders than the majority of us. there’s a reason why tech companies now look for people with experience in agentic tools
@StandUpForBarry@MatthewW_eacc@webmica you can critique it for being trained on and used for media but it’s become ubiquitous for productivity especially in swe. that’s just the reality unfortunately