All of ashtree's Comments + Replies

ashtree40

I also was trying to use Claude Code and it kept attempting to solve a problem by hard coding the output for n={1,2,3}, because that allows it to pass the tests, but it does not work in general, so this isn't exclusive to training. I tried multiple times with the same function and every time it couldn't implement it.

ashtree10

I am getting Forbidden 403 errors and invalid JSON parsing errors: started with '<html> <head> <...' while navigating the website. I am using eduroam and mobile data and it's occuring on both, so I don't think it's my issue. Where should I report this?

ashtree30

I don't know how useful that on arvix would be (I would suspect not very) but I will attempt to deliver a PDF by two days from now, although using Typst rather than LaTeX (there is a conversion tool, I do not know how well it works) due to familiarity.

Do you have any particular formatting opinions?

ashtree20

Did you see the XKCD? The chart there gives a good heuristic because most things you do that are worth optimizing are things you do at some interval.

I don't understand your second point. My guess at an interpretation is basically look up the optimal solution, but I don't think that makes sense with caching.

1daijin
I did see the XKCD and I agree haha, I just thought your phrasing implied 'optimize everything (indiscriminately)'. When I say caching I mean retaining intermediate results and tools if the cost to do so is near free.
Answer by ashtree40
  • If something takes longer than expected to get to an answer, think about how you could have seen the problem to be able to solve it faster to be able to solve it faster in the future[1].
    • As an extension, optimize everything. If something is slow and you do it often enough, doing it faster is worth it. Relevant XKCD.[2]
  • Flip a coin if you are struggling to decide between option in a situation where there are relatively low stakes. This exposes to you your gut instinct immediately, which is more than good enough most times, and it is far faster than logically
... (read more)
3daijin
disagree with the everything part of optimize everything. instead we need * a heuristic to determine if something is worth optimizing. i propose a back-of-envelope calculation of how much time you will spend on this in the future; or in practice I go 'have i done this three times?' * ways of doing things that are optimal in the first place: cache it if it's free to cache; write down the instructions before implementing.
ashtree30

Thank you for your advice, I donated $500. I expect to be a millionaire with high confidence (most situations where I am not are awful for humanity). "Owe" isn't how I would describe what I am doing; it's like any other donation to me. I want the best outcome for humanity.

ashtree10

An AI system replicating itself seems very unlikely because AI labs are hopefully and presumably protecting against that in particular. However, there are many other dangerous things an AI system could do that aren't self-replication and are often worse. It also seems that if that does happen, we are doomed since AI labs are trying their hardest to prevent that, and if they fail to prevent that, we have a self-replicating non-aligned AI, and so we are screwed.

ashtree10

I am reading Planecrash (https://www.projectlawful.com/boards/215), and it seems surprising that Keltham would not think of his mind being read after being informed of gods that purportedly can do so. This appears to be his biggest flaw so far that I can truly blame him for (lacking cultural understanding in many ways is easily excused by being an alien).

Second thought: what should someone going into the second semester of their second year of college for CS, math, and EE be doing? I have personally elected to continue to do college (it is paid for for me)... (read more)

5cata
I would be surprised if it were ethically important for you to donate that much. LW has made a pretty big difference to my life (e.g. my career, marriage, and a big chunk of my bank account are causally downstream of LW existing) and I estimated that there are probably something like $100m dollars worth of people for whom it was similarly impactful as me, and then a long tail of more people for whom it was somewhat less impactful, so I owed on the order of 1% of my net worth, such that if everyone like me who saw this fundraiser did the same then it would have enough money to thrive. So unless LW was really important to you or unless you are sure that you will be a millionaire in the future and you are just donating in advance, I don't think you owe $1000. But if you want to donate it, then do.
ashtree30

I do not use satellite view either, but my reason is mostly to do with the higher data usage (I'm on a pay-as-you-go plan). Even still, I don't think that there are many advantages to the satellite view especially considering the fact that roads can be obscured, and you still see the shapes of houses, traffic lights, and stop signs on Google Maps even without satellite view. Perhaps I am biased because I grew up in a place where there are trees everywhere, but it doesn't seem very useful to me.