longer.blue

I subscribe to both yours and Simon's blogs, and i can't really see this parody as good-faith. I respect your anti-AI stance, however, this post reads as someone who pre-maturely dismissed or is intentionally ignoring this increasingly viable strategy of SWE. Opus 4.6 or GPT-5.4 in the right harness can produce incomparably better results than the models of yesteryear. As a reader, I would appreciate a more informed and measured criticism on this topic instead of this parody.