"Columbia student suspended over interview cheating tool raises $5.3M to ‘cheat on everything’"
https://bsky.app/profile/hypervisible.bsky.social/post/3lne5zqaxyc2c
"Columbia student suspended over interview cheating tool raises $5.3M to ‘cheat on everything’"
https://bsky.app/profile/hypervisible.bsky.social/post/3lne5zqaxyc2c
But I’m not going to repeat what what I said about that bullshit here before already
Yeah, I didn't write at truly Rationalist length about it, but I did spend longer than was healthy.
An lesswrong:
countless articles that have ruined careers, stifled research, and brought entire fields of inquiry into undeserved disrepute.
uh-huh
An different lesswrong:
LLMs can provide reasonable fact-checks.
Christ on a futa dick
The closest I could find to an official statement is at the bottom here:
We are under advisement from counsel not to discuss any ongoing litigation in public at this time. We'll let you know the status of the current issues when we are able to. Your patience and understanding are most appreciated.
Further up the page, an IP editor says that another user is probably TW; the user in question has since been blocked for "ban evasion". Dunno what that's about.
Dan Olson finds that "AI overviews" are not as constant as the northern star.
The phrase “don’t eat things that are made of glass” is a metaphorical one. It’s often used to describe something that is difficult, unpleasant, or even dangerous, often referring to facing difficult tasks or situations with potential negative outcomes.
But also,
The phrase “don’t eat things made of glass” is a literal warning against ingesting glass, as it is not intended for consumption and can cause serious harm. Glass is a hard, non-organic material that can easily break and cause cuts, damage to the digestive tract, and other injuries if swallowed.
Olson says,
Fantastic technology, glad society spent a trillion dollars on this instead of sidewalks.
Kicking off the week's Stubsack with the evening's example of "why do I know who these people are":
Paul Graham approvingly mentions Jordan "Cremieux" Lasker as one who "has spoken out against both wokeness and authoritarianism".
Via the above search, here's some made-up bullshit:
https://en.wikipedia.org/wiki/Recursive_self-improvement
And there's a random LessWrong reference in the goddamn introduction here:
Ah! Further fucking around leads to a search query for all instances of lesswrong.com
in page source:
I got curious whether the Wikipedia article for Bayes' theorem was burdened by LessWrong spam. I don't see overt indications of that, but even so, I'm not too impressed.
For example:
P(B|A) is also a conditional probability: the probability of event B occurring given that A is true. It can also be interpreted as the likelihood of A given a fixed B because P(B|A) = L(A|B).
The line about "likelihood" doesn't explain anything. It just throws in a new word, which is confusing because the new word sounds like it should be synonymous with "probability", and then adds a new notation, which is just the old notation but backwards.
P(A) and P(B) are the probabilities of observing A and B respectively without any given conditions; they are known as the prior probability and marginal probability.
But both P(A) and P(B) are marginal probabilities; they're the marginals of the joint probability P(A,B).
The first citation is to one random guy's book that's just his presentation of his own "subjective logic" theory. And that reference was originally added to the article by (no prizes for guessing) the author himself, writing a whole section about his own work!
There are long stretches without citations, which I've been given to understand is frowned upon. On the other hand, one of the citations that does exist is to a random tutoring-help website whose "about us" page crashed Firefox on my phone. (I have been trying other browsers on my laptop, but not on mobile yet, due to finiteness of brain energy.)
To summarize that blog post and the three "Rationalist cases for Trump" that it points to: "We made up a Trump to like and a bunch of Democrats to get mad at".
The only one I've met is Carlo Rovelli, at the APS March Meeting in 2019. We work in adjacent topics, but I don't travel much. I know people who know him better, and I haven't heard stories of him being horrible, for whatever that's worth. He does come across as a bit of an eager self-promoter. I can easily imagine him accepting an invitation on the "a gig's a gig" principle.