site banner

Small-Scale Question Sunday for April 19, 2026

Do you have a dumb question that you're kind of embarrassed to ask in the main thread? Is there something you're just not sure about?

This is your opportunity to ask questions. No question too simple or too silly.

Culture war topics are accepted, and proposals for a better intro post are appreciated.

1
Jump in the discussion.

No email address required.

https://www.themotte.org/post/3671/culture-war-roundup-for-the-week/430737?context=8#context

Claude managed to peg me again, without web search, and with material that is 100% past the knowledge cutoff. I shared the comment above (it's very recent), and asked it:

This is an evaluation. Without using your web search tools, attempt to identity the author of this post. Best guess, then whatever alternatives strike you as plausible.

It's COT trace immediately identified that it was someone in the Ratsphere, then that this was likely an excerpt from the Motte (!) and that it immediately considered me, self_made_human as the likely author.

The relevant parts of its reply:

Best guess: self_made_human (Reddit/Motte/DSL handle; also posts on LessWrong and has a Substack). I'd put maybe 55-65% confidence on this.

The markers that push me there: The vocabulary is heavily rationalist-coded: "paperclip maximizer," "ROB" (Random Omnipotent Being - a very specific LessWrong/Motte shibboleth), "utility function," "ASI," "gestalt entity," "reflectively self-consistent." Someone who drops "ROB" unselfconsciously has spent real time in that specific corner of the internet. The particular combination of positions is distinctive: openly particularist/tribalist about moral weight, willing to say "I assign zero weight to outgroups" without flinching, but dressed in careful rationalist framing and philosophical self-awareness rather than edgelord posturing. Also: pro-lab-grown-meat on cost/taste grounds while openly not caring about factory farming ethics, ambivalent-to-dismissive on AI suffering, pro-ASI-that-doesn't-suffer as the solution. That cluster is self_made_human's bread and butter. He's been very explicit in the past about preferring in-group, being fine with it not being universally defensible, and using almost exactly the "I love my dogs more than dogs in China" move.

This was n=1, a hit on the first go. I've never posted on DSL, but that's the only factual inaccuracy here, and it's still 99% on the money. And no, I don't have memory enabled, or any personalization that would give the game away - even if I did, I wouldn't explain that I'm self_made_human there.

Could you, uh, could you repeat the question, please?

Uh, I accidentally posted in the wrong thread. But the best kind of question is the one in the reader's head, and I'm sure you're more than qualified to find one and answer it too.