Replying to
@jzb@hachyderm.io Because there's not just random people with an LLM, there's people that know what they are doing, yet find it easier to have an agent do the manual task of writing the code. As I said, there are people that have difficulty doing the coding part, but the testing and the rest they can do without a problem. Its more nuanced than just random people using LLMs to generate code and then make PRs. The people at KDE did it right, no promoting LLMs, just good quality code regardless how it was written. @dansup@mastodon.social