Ugh, a lot of tunnel vision in here. Good open source projects often last years or decades and are holding most of the internet alive today. They used 0 LLMs and are starting to use it today with dubious results. Many professionals ghat used it in programming reported an initial boost in production followed by an incedible boost in deal breaking bugs and higher fixing times: turns out that if you don’t directly write code debugging is hell and all the gains vaporize when you need to manually verify every line again. That said, mental atrophy is also worth researching. All in all, whoever uses it in fun projects and didn’t formally study code is always the most vocal about LLMs and those who did study and are forced to use it by their boss always describe it as bad. Wonder why
I’m italian and I agree with you