pmichaud

9 ms·

61.
by pmichaud 4y ago
https://web.archive.org/web/20230307024151/http://ascii.text...
62.
by pmichaud 4y ago
This was a really good read! It strikes at the heart of some frustration I feel when people get smug about "the science" as if they have checked anything for themselves. Seriously: be less confident.
63.
by pmichaud 4y ago
I have no idea what this abstract means. Does someone have a nonpaywalled version?
64.
by pmichaud 4y ago
In finance they use it because "M" is the roman numeral for 1000. MM is 1000*1000.
65.
by pmichaud 4y ago
I have seen the same thing, but I think the cause is more that maybe the overwhelming majority of people can’t meaningfully do abstraction? Like they blow a gasket between conceiving of individual actions and beliefs vs policies related to
66.
by pmichaud 4y ago
As the sibling comment notes, there is in fact markup for this, and the results are actually pretty great.
67.
by pmichaud 4y ago
I think I get the gist of your point which I would handwavily summarize as “there is no there there,” which I agree with. But I’m thinking that in terms of functionality, there is no real distinction here. Like If I train a model to do vi
68.
by pmichaud 4y ago
What tribes did with people who couldn’t pull their weight is an empirical question that varies per tribe per time. But, I think you’re essentially correct, and far from being a gotcha, your observation that going out on the vision quest di
69.
by pmichaud 4y ago
I kind of feel like _all it can do_ is seek new abstractions. It’s clustering the data into what you might think of as “concepts,” even though we haven’t made all that much progress on interrogating those concepts directly. I just think clu
70.
by pmichaud 4y ago
I have a hunch that societies generate strong taboos specifically in response to things that are widely appealing to individuals but that are somehow considered harmful to the whole.
71.
by pmichaud 4y ago
I think the reason, other than easier implementation, is there is a rule/heuristic that you shouldn't move controls around. So having the controls work one way in full screen and a slightly different way outside of the fullscreen
72.
by pmichaud 4y ago
Yeah I dunno. I mostly don't interact with LW, I mostly interact with people who are part of the in-person core of the community, so maybe the people online just have less realistic hopes than the core does.
73.
by pmichaud 4y ago
What I’d actually like to see is more “dynamic” music. the format of a song used to be limited like a book is limited, but then we invented hypertext. Paintings are limited, but then we invented video games. Why can’t music have various int
74.
by pmichaud 4y ago
I’m not signed up, but I know many people who are, and none of them assume any of these things. No one “expects” it to work, it’s just that they figure a long shot is better than no shot. The people I’ve talked to about it put their chances
75.
by pmichaud 4y ago
I don't think I agree that this is just semantics. I looked into it enough to realize it was just a simulation, but the headlines were all about how they had created a wormhole, very obviously because a headline like "researchers
76.
by pmichaud 4y ago
Ideally the equilibrium would be that enough people in the group can't be bullied that it extinguishes the bullying behavior, or bullies get ganged up on.
77.
by pmichaud 4y ago
My guess is that filtering is downstream of prediction, because filtering is making a prediction about which input will be useful, then acting on that prediction.
78.
by pmichaud 4y ago
To be fair to those people, they probably, by and large, have no idea what to actually do. Starting from nothing, changing your diet to something healthy and learning how to be active at all is kind of huge, like maybe a university course o
79.
by pmichaud 4y ago
What signs do you see that we’re headed toward a new, stable equilibrium, instead of just catastrophically destabilizing?
80.
by pmichaud 4y ago
I’m having trouble reading this comment as sincere, even though I think it probably is. You’re not trolling?
81.
by pmichaud 4y ago
I guess the intuition is that it works like a market to converge on truth, but of course there’s no skin in the game and the metrics are all wrong for actual market dynamics.
82.
by pmichaud 4y ago
Wood is thicker, denser, and has more structure than paper. My guess is that paper is a closer analogy in this case.
83.
by pmichaud 4y ago
This is essentially correct, yes.
84.
by pmichaud 4y ago
Is it a faux pas? I hear this sort of thing constantly, eg. your post?
85.
by pmichaud 4y ago
Re: your edit, I think the real answer is that the correct translation is fully context dependent, and the people who want it strict want to basically force/precompute the context, vs the people who want it loose want the context itsel
86.
by pmichaud 4y ago
Sounds like totally normal Bay Area behavior :)
87.
by pmichaud 4y ago
The logic isn't that student loans are a special thing that shouldn't have to be paid back. The logic is that under normal circumstances the lenders has an incentive to evaluate the probability that the borrower will actually pay
88.
by pmichaud 4y ago
This is essentially the model that we have with the FDA, right? And we have a bunch of problems with that too. I don't think there's a simple solution that just makes all scenarios better/easier/safer/whatever.
89.
by pmichaud 4y ago
This seems like a useless place to optimize. Like, what is the total pollution output of all Hummers or similar vehicles? A rounding error, I'm guessing. The only way to make dents is with systemic changes like for entire trucking indu
90.
by pmichaud 4y ago
I do think it's related to age, and appears "out of nowhere." A lot of people describe it as "waking up one morning, and..." I don't know the mechanism, but sure, it might be hormonal.

More ›