Counting calories and abiding by your commitment to eat fewer of them is hard, hard enough that while it worked for me I can't just recommend it to everyone. Like yeah no shit eat less food. It's really hard to admit that there are limits to what people can do with willpower alone, especially if you live in a subculture where you think that being galaxy brained should allow you to do literally anything. There must be some reason, any reason, besides people not being fully 1000% in control of their actions.
Yeah, it's definitely really hard. The hard part is not "knowing that eating less food will make you lose weight," it's actually doing the thing without suffering from willpower failure. But, even given that, Yudkowsky seems to be arguing here that eating less calories won't make you lose less weight, because such a simplistic model can't possibly be true (analogizing it to the silly idea that eating less mass will make you lose weight.)
However, uh, his conclusion does contradict empirical reality. For most people, this would be a sign that they should reconsider their chain of logic, but I guess for him it is instead a sign that empirical reality is incorrect.
First you take some food in. Then you apply the Krebs cycle. Then you suddenly learn a lot more about organic chemistry. Then things start getting into the funky. What’s hard? :sarcmark:
But you already lost Yud right after step 1, so you’ll have an edge up either way
Not like he's ever been a competent physicist but he has always had the attitude of STEM chauvinism down pat.
Everything obeys the laws of math and logic, right? Let's just Russel&Whitehead everything out of Zermelo-Fraenkel set theory. Or better yet, something I personally came up with based on vibes and NIH.
This also informs his anti-empiricism. Why bother actually looking or listenging when you can just work everything out from a handful of assumptions.
I had a coworker once who was vehemently opposed to lights on the bike path because, he showed us through a chain of impeccable logic, being able to see victims would make it easier to assault people.
But what about all these studies, I asked, showing empirically that lighting pedestrian/bikeways leads to less assault?
Nonsense data, he informed me. Correlation not causation.
Nothing at all to do with him living behind the path & not wanting lights. Nope. Just unassailable logic.
Started somewhere in 1995 when he discovered the internet.
But yeah it reads like he has just discovered weight loss stories and is first principles reasoning out all the 'hypocrisy' in an 'I just learned logic from wikipedia' guy style. (I found an extreme focus on people being hypocritical or having flaws in their logical systems if you drag out the logic to extremes to be a very right libertarian thing, so that is worrying for the younger people in the Rationalist community ;) ).
It gets amusing as he talks about calories a lot but never really points to the whole 'how are calories actually measured' thing (they burn it and see how much it heats up water (or add up the various macronutrients)). As the Rationalists have a huge blindspot in not thinking hard about faulty measuring systems. He is this close to arguing that cows living on grass (pretty low in kcal (even zero effective ones for humans iirc)) violates the laws of physics.
(He prob is also losing his mind due to his lifework exploding all around him. He is starting to see (AGI doom denying) conspiracies, and even Musk (who they thought would be a techno savior) is building AI, and his solution to solving the AGI alignment problem is 'just give it human friendly rules' and 'weak grip strength'. Showing he understood nothing from the whole AGI alignment problem).
Is this the "flying saucer fails to land" moment for him? AIs (large language models, ai-generated images, etc etc) are now within the experience or understanding of more and more people, and he can't just make stuff up about it anymore?
Seems to be a/b testing a new posting style aimed at a... less discerning audience of technoilliterate anti-wokists, I think, while claiming that people missed the joke/deep philosophical point he was illustrating without actually meaning what he wrote every time a post falls embarrassingly flat.
Once he gets that nobody outside the handful of rat forums is actually bothering with his bizarre wall of texts rants, i think blue collar vlogging behind the wheel yud night be the logical next step.
Academic prose is not known for its scintillating qualities, but reading meta-analyses and reviews about calorie restriction (for example) is easier than reading a Yud tweet.
I mean he's not wrong to criticise the super simple calories in calories out, or even how food calories are determined, but like nobody thinks that model is true except reply guys on reddit.
Our understanding of food has moved on, you don't have to be such a dork about saying "hey if you think this maybe read some more recent literature".
fuck me I’m getting flashbacks to the worst posts I’ve seen on the internet, all thanks to Reddit’s weird hangups: calorie reply guys, the fucking pitbull thing, vegans who read like fictional vegans made up by people who hate vegans, all the folks trying to convince you you’re abusing your pets
I still can't bring myself to believe that straw vegans exist. I know they *do*, but they don't resemble any real vegans I've ever met, so in my mind the straw vegans *must* be fictional vegans made up by people who hate vegans.
All vegans are awesome. Even annoying ones. It's wild how much flak people throw out. Like I wouldn't criticise an annoying peace activist in front of war hawks because no matter what they're doing they're way less disturbing than the war hawks.