Case in point, the recent Hank Green incident. That was an own-goal of being too honest about LLM use and I suspect the industry learning will be "don't talk about the process." Larian Studios got in hot water earlier over a similar "we may use AI in preproduction" incident.

Cheating isn't the right word here. Crediting LLMs is about giving everyone involved accurate information which can be used for later decision making (e.g. if something unexpected happens as a result of a LLM-ism).

I can confidently say I do not have an encyclopedic knowledge of blog posts and who wrote them I may or may not have read over a year ago. I made that comment on that post because it's weird to quote yourself, I don't think I actually read the post.

I have no grudges against any particular Hacker News user and I typically don't pay attention to who writes a given post/comment unless it's out-of-pocket.

Technical blog posts don't need to follow AP Style but this post reads like LinkedIn broetry.

Also you can't do the "this used to be an actual place where posts meant something" bit when your account is 3 months old.

(the fact that it's currently at #3 is a bit weird and I've noticed some things weird recently with the HN algorithm but that's off topic)

I've been fine with the G502 X Lightspeed (which should avoid said double-click issue due to optical switches) and some more efficient button use to compensate for the fewer side buttons.

FlashAttention was a hell of a drug.