TechTakes

2648 readers

49 users here now

Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.

This is not debate club. Unless it’s amusing debate.

For actually-good tech, you want our NotAwfulTech community

founded 3 years ago

MODERATORS

dgerard@awful.systems

Anthropic watermarks AI output — and the AI bros yell (pivot-to-ai.com)

submitted 6 days ago by dgerard@awful.systems to c/techtakes@awful.systems

9 comments fedilink hide all child comments

how DARE you catch me out

https://www.youtube.com/watch?v=sCiDuw5CzAY&list=UU9rJrMVgcXTfa8xuMnbhAEA - video
https://pivottoai.libsyn.com/20260817-anthropic-watermarks-ai-output-and-the-ai-bros-yell - podcast

time: 7 min 46 sec

you are viewing a single comment's thread
view the rest of the comments

[–] dgerard@awful.systems 4 points 5 days ago* (last edited 5 days ago)

yeah. The word-choice watermarking is an interesting trick for a paper, needs quite a chunk of text, will have a ghastly rate of both false positives and false negatives in real world fraud detection, and is evidently good enough for Gemini not to drink too much its own piss but not much else.