EDIT 2: To be clear, I was doing the exercise Ben recommended at the start of the post: predicting what I think the worst, credible info he could’ve found was. And this was a quickly written sketch of my off-the-cuff predictions. To be clear, the upper bounds I give are loose bounds and I lumped some really bad stuff with not-so-bad stuff. The me who wrote this comment a couple days ago would’ve been shocked if anyone in Non-linear had done the worst stuff in this list. But admittedly somewhat less shocked than the median LW user, I think. Milder forms of misconduct, like corruption and contract breaking, would’ve been most of the 1⁄100 chance I mention below.
A quick sketch of my estimates about how bad Ben’s claims are:
I think there’s like a ~1/1000-1/100 chance for each person amongst Non-linear to have been a rapist.
Maybe a 1/100-1/10 chance of other bad things having been done by people in Non-linear. Much/most of that probability mass is in miscelaneous forms of corruption. But I’d put maybe ~1/100 chance of any of corruption, blackmail, embezzlement, breaking major contracts, risking people’s lives, beating someone up, and admittedly very unlikely, but possibly seriously physically injuring/killing someone.
Minor corruption or failures in rationality happening in the organization happening as a whole somewhere seem quite likely?
Edit: To be clear, this isn’t because I’ve heard anything wrong about Non-linear previously. I’ve just updated hard enough on evidence of malfeasance amongst rats/EAs that I think that rat orgs are pretty average in how uncorrupt they are, and rats are pretty average in how moral they are, given their demographics. It is hard to say whether the initial part of this post, and some tangential discussion of it, affected my estimates.
I think the downvotes are coming because people don’t realize you’re doing the exercise at the start of the post, and rather think that you’re making these claims after having read the rest of the post. I don’t think you should lose karma for that, so I’m upvoting; but you may want to state at the top that’s what you’re doing.
Ex ante, it is obvious I should’ve mentioned that. But I just saw a bunch of comments making these guesses, and I thought my edit made it clear that these estimates weren’t based on the post. Also, I really should’ve been more clear about the low likelihood of the really bad stuff.
Even now I would like it if you added an edit at the start to make it clearer what you’re doing! Before reading the replying comment and realising the context, I was mildly shocked by such potentially inflammatory speculation and downvoted.
EDIT 2: To be clear, I was doing the exercise Ben recommended at the start of the post: predicting what I think the worst, credible info he could’ve found was. And this was a quickly written sketch of my off-the-cuff predictions. To be clear, the upper bounds I give are loose bounds and I lumped some really bad stuff with not-so-bad stuff. The me who wrote this comment a couple days ago would’ve been shocked if anyone in Non-linear had done the worst stuff in this list. But admittedly somewhat less shocked than the median LW user, I think. Milder forms of misconduct, like corruption and contract breaking, would’ve been most of the 1⁄100 chance I mention below.
A quick sketch of my estimates about how bad Ben’s claims are:
I think there’s like a ~1/1000-1/100 chance for each person amongst Non-linear to have been a rapist.
Maybe a 1/100-1/10 chance of other bad things having been done by people in Non-linear. Much/most of that probability mass is in miscelaneous forms of corruption. But I’d put maybe ~1/100 chance of any of corruption, blackmail, embezzlement, breaking major contracts, risking people’s lives, beating someone up, and admittedly very unlikely, but possibly seriously physically injuring/killing someone.
Minor corruption or failures in rationality happening in the organization happening as a whole somewhere seem quite likely?
Edit: To be clear, this isn’t because I’ve heard anything wrong about Non-linear previously. I’ve just updated hard enough on evidence of malfeasance amongst rats/EAs that I think that rat orgs are pretty average in how uncorrupt they are, and rats are pretty average in how moral they are, given their demographics. It is hard to say whether the initial part of this post, and some tangential discussion of it, affected my estimates.
I think the downvotes are coming because people don’t realize you’re doing the exercise at the start of the post, and rather think that you’re making these claims after having read the rest of the post. I don’t think you should lose karma for that, so I’m upvoting; but you may want to state at the top that’s what you’re doing.
Ex ante, it is obvious I should’ve mentioned that. But I just saw a bunch of comments making these guesses, and I thought my edit made it clear that these estimates weren’t based on the post. Also, I really should’ve been more clear about the low likelihood of the really bad stuff.
Even now I would like it if you added an edit at the start to make it clearer what you’re doing! Before reading the replying comment and realising the context, I was mildly shocked by such potentially inflammatory speculation and downvoted.