niplav.site

No description provided.

“showing that you care” type signalling depends on the sunk cost fallacy; or, there was selection pressure against the sunk cost fallacy (modulo reputation effects)

⤋ Read More

Let A be the set of AI alignment complete problems. What is the problem a /∈ A so that solving a maximally delays the development of AGI? What is a, weighted by how hard/desirable it is?

⤋ Read More

I think I understand 5-10 now. It basically says “Let’s say I jumped out of the airplane. I would only do that if I had a parachute. So, any version of me that jumps out of the airplane has a parachute. Therefore, if I jump out of the airplane, I’ll have a parachute. Let’s jump out of the airplane!”

⤋ Read More

probably similar to solomonoff, just in the number of variables, and without the computability aspect (since we know that prop-formulas can be evaluated in exponential time)

⤋ Read More

major difference between updating on evidence vs. morals: we know that a bayesian is converging (monotonically?) towards truth with each piece of evidence, while with moral progress, it seems likely that we are not bounded in how wrong our updates can be (even if we grant that the arc bends towards justice in the end). we might need to escape moral local maxima. therefore, it seems good to preserve option value.

⤋ Read More

I can often distinguish sympathetic mockery from sneering mockery, even with relatively litlte context. I wonder why – what am I picking up on? What is the distinct way of making fun of someone you actually like vs. don’t like?

⤋ Read More

A: “X is number”. B: “Thats not true, X is not number”. If B doesnt provide an alternative number (or an explanation of why a number is not applicable here), Im often more inclined towards believing A.

⤋ Read More

2x3: {lives in social/physical reality}×{views things generally as positive/zero/negative sum}. To be honest, I think there’s relatively few people in social positive sum reality frames.

⤋ Read More

if you’re weird, spending time around non-weird people is not going to make you more normal. but if you spend time with marginally less weird people, you’re probably going to become less weird.

⤋ Read More

@movq@www.uninformativ.de “Random thought: Would be great if you could do for i in ...; do something "$i" & done ; wait in a Shell script, but with the Shell only spawning one process per CPU.” -> Interesting which annoyances stay in the back of the head – I’d never articulated this, but it’s absolutely true that this would be great.

⤋ Read More