nathan_young

Seems like hackernews should use a bridging algorithm for flagging.

Stop being rude. There were multiple OAI hacking incidents. Irregular's evals were involved in some but not the HuggingFace hack. Hence Aesthesia is both accurate and precise.

If there is interest here I may add other streams. Would you sign up to such a site?

Happy to answer questions as to the sources used (Kalshi and Metaculus) and the conditional probabilities assigned: - P(bird flu as bad as covid | 10,000 US cases) = .5 - P(bird flu as bad as covid | CDC travel advisory) = .1

They came out of Manifold. Though I acknowledge that doesn't really answer your question.

Ad what % is it crazy shit? I'm at about 7% of human disempowerment/exctinction, so it probably won't happen. Am I crazy?

According to a study of 700 ML researchers (not AI safety) the median chance that that the long-run effect of advanced AI on humanity will be “extremely bad (e.g., human extinction)" is 5%. And 48% of respondents gave an answer of 10% or more.

Even if this is 2 orders of magnitude too high it's worth a lot of our time. But these are probably mainly normy researchers not AI safety-ists. https://aiimpacts.org/2022-expert-survey-on-progress-in-ai/#...

If you see any tech tools or data please add them here:

Coronavirustechhandbook.com

A lot of these tools have very little money or support. Some of them have had 1m+ views last election.

If something interests you, you could offer to help the developers involved.

(disclosure: I am involved in the project)

Please add it in.

We (Newspeak House) would love to see one created. If you know people in that space, put up some tech and share it. Feel free to use our template if you credit us.

Feel free to edit text to note if tools are biased/add tools with different biases.

(disclosure, I helped initiate the doc)