I know you're being downvoted, because people are tired of hearing "this is AI".

But I think your comment "this is AI" is warranted. This content, explaining something of cultural, personal significance (instead of the typical technical explainer) should really be written by a human.

astrobe_ 3 hours ago | parent | next [-]

Possible false positive because of the text was translated from Japanese to English with AI?

I don't see the point of starting a newsletter written by AI.. But I don't understand the point of cheating with a computer on lichess.org either.

	▲	wavemode an hour ago \| parent \| next [-]
		> Possible false positive because of the text was translated from Japanese to English with AI? People say this a lot, but I've never seen Claude or any other LLM insert clichéd LLM-isms into translated text. There's obviously a degree of interpretation and "smoothing" that goes into translation, but it doesn't rise to the level of making the entire thing sound like an AI's voice. This blog post was prompted, not translated.
	▲	StilesCrisis 28 minutes ago \| parent \| prev \| next [-]
		A Japanese blog wouldn't discuss this. It'd be like an American blog with an article about "what does thank you mean." We already know.
	▲	twister2920 an hour ago \| parent \| prev [-]
		that's a true positive

▲

asgraham 4 hours ago | parent | prev [-]

How good are these detectors? (Serious question— I’ve never had a reason to use one)

	▲	no_multitudes an hour ago \| parent [-]
		Pangram's actual classifier model is really strong. It has an incredibly low false positive rate. Whenever people complain about it, I ask them to provide verifiable non-adversarial long-form false positives, and nobody has done so yet. False negatives are more common (as they should be -- it's better to let an AI article through than to false accuse someone.) Their frontend is often misleading, though, for these reasons: - 100% means that it thinks 100% of the article is AI, not that it's 100% confident the article is AI. It's very easy for people to get this confused. - It is much less accurate on shorter text, and doesn't advertise this fact clearly enough. - It works on chunks of text (I think they're like 100 tokens) and can't identify within that chunk which parts are AI and which parts are human.) This should be more clear in the UI, I think.