domain:twitter.com
gattsuru
3hr ago
I get that Trump wanted to shut down "woke" research, but he could have done that without cutting overall funding (just mandate that the NIH can't fund transgender research, shutdown the diversity grants, etc.).
Could you? One of the morbid bits to this saga has been how often people have pointed to what they saw as clearly misaimed anti-DEI efforts that must have been motivated by an LLM or a bad grep, and then oops.
Yes, there are research areas with neither blatant political abuse or outright woke goals. But the people who want to do the woke research can, as warned, lie: there’s far less signal than anyone thinks to a research’s quality from how sober the grant application.
odd_primes
3hr ago
Saagar from Breaking Points says he has a copy of the arrest documents and will release them today.
thejdizzler
4hr ago
Definitely not. Unlike in a lot of other areas of life there is actually quite a bit of transfer between different sports, especially in the endurance world. Alan Cousins actually (surprisingly to me) found that biking was 2/3 as effective as running at increasing running Vo2 max. This is insane and runs contrary to what I had been taught about specificity growing up, but makes sense from what we know about elite training (Cole Hocker who won the gold in the 1500 in Paris apparently bikes a shit ton), and from my own training (I am fastest at running when I supplement with monster bike time). Here's the Couzens post if you're interested: https://alancouzens.com/blog/specificity.html
Maybe not directly relevant to your post, but at least a tangent. Fitness from random things generally does carry over pretty well because it develops your core cardiovascular and muscular systems. There doesn't seem to be an equivalent "core" system with regards to intellectual pursuits so it's not surprising we don't see much transfer there.
Do we condemn Kolmogorov?
I don't condemn enemy conscripts. "The enemy" is not necessarily synonymous with "evil", and that's something lots of people have forgotten (the Nazis are a foundation of how the American Empire justifies its right to rule to itself, so it's kind of unavoidable)- if my enemy forces all of its constituent parts to, for example, wear a blue shirt or die, I don't blame anyone for putting the blue shirt on [whether or not they share all of my enemy's goals is irrelevant].
Yet, I don't think friendly forces are evil for killing them either- even in an environment where the enemy has intentionally frustrated identification of those who cause the enemy's cause (those who would rather die before ceasing to be the enemy), and those who would abandon those principles to not be dead (this includes those who only joined for the meals).
It is not, and cannot be, the enemy's fault that circumstances forced your uniform upon you; your only hope is that your own side advances its interests in such a way that your enemies do not decide to violently destroy you if and when they obtain the power to do so.
He does not necessarily deserve the consequences of being an enemy (contra traditionalist thought, where he does), but at the same time it is not immoral to destroy enemies (contra progressive thought, where it is), so I guess it depends on what you actually mean by "condemn".
phailyoor
4hr ago
I'm a bit more sympathetic to Tao: he lives and works in a milieu where not signing that letter would have made many of his colleagues and students (maybe even his wife) shun him;
"Letter? What letter? Oh there was an email? I must have missed it. Can you send it again I'll definitely put it at the top of my queue for sure."
Academics are absolute masters at ghosting and dodging, as everyone who has set foot in a school can attest to. And I'm 100% confident that there are plenty of other UCLA professors who didn't sign the letter. So given that he didn't just ignore it, he's fully responsible for the consequences of signing that letter.
Edit: I'm not going to bother checking the entire list, but the very first 2 professors in the math department aren't on the letter: https://web.archive.org/web/20200807214114/https://www.math.ucla.edu/people/ladder
TitaniumButterfly
4hr ago
Very well-said. The thing about inviting cleansing fire is that it's not exactly discriminate.
phailyoor
4hr ago
He did good finding that letter but Trace is definitely a huge tool for not posting the actual link - I believe this is correct: http://atripati.bol.ucla.edu/May2020AntiRacismLetter.htm
SnapDragon
4hr ago
It was at least somewhat justified by the bullshit tech the aliens had. (Which very conveniently could completely control all scientific experiments but not, you know, actually KILL anyone.)
The real problem with The Dark Forest (spoiler alert) was the concept that all of humanity, working for more than a century on a problem with existential stakes, failed to come up with a theory that, uh, most people interested in cosmology already knew about in the 70s as a potential answer to the Fermi paradox. (Also, the deterrent threat at the end doesn't even really work because it would send a message out only in the plane of the ecliptic. Sigh. I wouldn't mind the bad science so much if it weren't wearing the skinsuit of Hard Sci-Fi.)
EverythingIsFine
Well, is eventually fine
4hr ago
It only takes about 5 bucks and a few minutes to use the API with a bring-your-own-key chat interface to pay as you go for Sonnet 4.
TitaniumButterfly
4hr ago
I'd appreciate some followup on Ride the Tiger. Never quite got into it and thought a (good) intro might be helpful.
What you perceive as "goofy" is in my mind more like "optimized for achieving balance across multiple domains of fitness."
Eh, I was thinking, like, shadowboxing with dumbbells, or anything involving Bosu balls or squishy foam mats or tsunami bars, for instance, none of which I would consider simple but balanced. Though, granted, it's not like I have a video montage of top MMA guys doing that stuff.
When you say you are suspicious of general fitness, are you saying such a property doesn't exist, that it's impossible to describe, or that it never matters to anyone?
I would accept either "doesn't exist" or "is impossible to meaningfully describe" as a characterization of my views, here's my reasoning:
optimizing fitness for a given activity A produces different levels of fitness for B and C;
This is of course correct, but I think that people's actual selection of A, B, C, ..., ultimately boils down to some.combination of the following:
-"idk it just sounds cool", great, awesome, that's pretty much what it comes down to for me as well, but I don't think you can get from this to meaningful claims about generality.
-muh fizeek, to be answered by a dismissive Bronx cheer
-fighting/soldiering/moving house/farming/etc from someone who's not actually doing any of those things and has no plans to start, ditto
--fighting/soldiering/moving house/farming/etc from someone who is actually doing one of those things, but then you're just doing task-specific s&c, and it's not going to matter much in comparison to specific practice anyway.
Basically, I don't think there's a principled way to select a truly general A, B, C.
On a purely autobiographical level, I experienced noticeably better carryover to manual labor in the woods from training like a dentist with a half Ironman coming up than I did from various well-regarded "tactical" training systems. I suppose this isn't a terribly widespread experience, but then again I don't know how many people have tried both, and it certainly made me more skeptical of the idea that I had to think about some kind of balance or generality in my training for it to carry over to real-world tasks.
TitaniumButterfly
4hr ago
It is definitely very 60s in its view of sexuality.
There's a lot of deep Heinlein no one talks about. The Door Into Summer flirts with some odd subjects, and Glory Road gets kind of out there, but probably nothing tops Farnham's Freehold.
the_others
4hr ago
If you were part of the Trump administration, how would you punish academics for their woke excesses without negatively impacting useful research? The federal government does not directly control how universities manage their own affairs and any penalties assessed on the universities as a whole can be cast as damaging research in some way.
The only thing that I can think of is some sort of rule like "any university that violates XYZ policy automatically becomes federal property", which would allow the federal government to directly fire and hire, but nationalizing the universities comes with a million other problems.
TitaniumButterfly
4hr ago
I'm feeling called out, but never actually said it was a grammar mistake.
TitaniumButterfly
4hr ago
This friend speaks my mind.
TitaniumButterfly
5hr ago
(((Scottish)))
The_Nybbler
If you win the rat race you're still a rat. But you're also still a winner.
5hr ago
Do we condemn Kolmogorov?
Scott doesn't. I do.
TitaniumButterfly
5hr ago
Wow, that really sucks.
The_Nybbler
If you win the rat race you're still a rat. But you're also still a winner.
5hr ago
I get that Trump wanted to shut down "woke" research, but he could have done that without cutting overall funding (just mandate that the NIH can't fund transgender research, shutdown the diversity grants, etc.).
The problem is that there isn't only some single class of "diversity grants". Every grant has some sort of DEI stuff written into it, including the main IPAM grant. It's a Gordian knot tying research to DEI, and there's only one way to deal with those.
TitaniumButterfly
5hr ago
Some major realizations about culture when I was a teen reading The Diamond Age on lunch at the job I had then. Just looking at the cover takes me right back even though I've read that book at least ten times now.
birb_cromble
5hr ago
That makes more sense. Thank you.
hydroacetylene
5hr ago
Scientists are born subjects.
If the delusional fever dreams of democrat true believer karens come true and a Christian theocracy rises to power, the scientific establishment will simply publish appendixes to their papers reconciling them to the current state of creation research. Woke is the same damn thing. Trump should be demanding they include 'murica, fuck yeah! loyalty pledges instead of yeeting them for kowtowing.
thejdizzler
5hr ago
Sorry I think my response was a bit confusing because I don't want to pin the blame solely on Trump for this. Universities have played with fire for a long time and somehow seem surprised to be getting burnt. I just lament that the administration seems to be cutting down the tree rather than pruning some of the worst branches. We can punish woke without destroying the research apparatus.
Why don't we try and look into this? People have tried to estimate OpenAI margins on inference and they come away with strong margins of 30, 55, 75%. We don't live in a total vacuum of information. When trying to work out their margins on inference, I base my opinion on the general established consensus of their margins.
The demand for inference is rising, Openrouter records that demand for tokens rose about 30x in the last year as AI improves. Grow big enough and the margin on inference will outweigh the costs.
It's effectively free, they're 'selling' it for $1 per agency for a whole year. OpenAI is doing the same thing. Why are you trying to correct me on something you won't even check?
There is a significant difference between making a loss as you expand your business rapidly and try to secure a strong position in an emerging market and 'subsidized by 1-2 orders of magnitude'. No evidence has been supplied for the latter case and it's unbelievable.
Amazon wasn't making a profit because they were continuously expanding and investing in their retail business, not because the actual business was unprofitable. Investors were happy to tolerate them not making profits because they were growing. Uber wasn't making a profit but there were no 10x subsidies. We can see this immediately in how taxis weren't costing $20 while Uber was costing $2 for the same trip.
More options
Context Copy link