|
|
|
|
|
|
by Flere-Imsaho
212 days ago
|
|
|
You've been downvoted, but this is exactly how I feel as well. There is an element of denial amongst the HN crowd, commonly saying stuff like "Well I tried Claude code and it produced garbage". Any task with a tight "write->test->repeat" loop is going to get AI-trained into oblivion, and we've only 4 years into this LLM disruption. Personally, I'm training up on: infrastructure, systems administration, security and software-architecture - because these are harder to train on given a longer "write->test->repeat" cycle, although I'm not in denial that they too will be disrupted. Writing code for any problem - especially Leet-code style problems - is going to be solved by AI eventually. Don't be left behind. |
|
|