Comment by Aurornis | Hacker News Reader

Comment by Aurornis | Hacker News Reader

Parent Full thread

> But there also is an arrogance that "we are the ones thinking about this rationally, and we will figure this out". As if people hadn't been thinking and discussing and (verbally and literally) fighting over all sorts of adjacent and similar topics in philosophy and sociology and anthropology and ... clubs and seminars forever

This is a feature, not a bug, for writers who hold an opinion on something and want to rationalize it.

So many of the rationalist posts I've read through the years come from someone who has an opinion or gut feeling about something, but they want it to be seen as something more rigorous. The "first principles" writing style is a license to throw out the existing research on the topic, including contradictory evidence, and construct an all new scaffold around their opinion that makes it look more valid.

I use the "SlimeTimeMoldTime - A Chemical Hunger" blog series as an example because it was so widely shared and endorsed in the rationalist community: https://slimemoldtimemold.com/2021/07/07/a-chemical-hunger-p... It even received a financial grant from Scott Alexander of Astral Codex Ten

Actual experts were discrediting the series from the first blog post and explaining all of the author's errors, but the community soldiered on with it anyway, eventually making the belief that lithium in the water supply was causing the obesity epidemic into a meme within the rationalist community. There's no evidence supporting this and countless take-downs of how the author misinterpreted or cherry-picked data, but because it was written with the rationalist style and given the implicit blessing of a rationalist figurehead it was adopted as ground truth by many for years. People have been waking up to issues with the series for a while now, but at the time it was remarkable how quickly the idea spread as if it was a true, novel discovery.

I don't read HN or LW all that often, but FWIW I actually learned about SlimeMoldTimeMold's "A Chemical Hunger" series from HN and then read its most famous takedown from LessWrong: https://www.lesswrong.com/posts/7iAABhWpcGeP5e6SB/it-s-proba... (I don't remember any detailed takedowns of SlimeMoldTimeMold coming before that article, but maybe there are).

I think that SlimeMoldTimeMold's rise and fall was actually a pretty big point in favor of the "rationalist community".

> I think that SlimeMoldTimeMold's rise and fall was actually a pretty big point in favor of the "rationalist community".

That feels like revisionist history to me. It rose to fame in LessWrong and SlateStarCodex, was promoted by Yudkowski, and proliferated for about a year and half before the takedowns finally got traction.

While it was the topic du jour in the rationalist spaces it was very difficult to argue against. I vividly remember how hard it was to convince anyone that SMTM wasn't a good source at the time, because so many people saw Yudkowski endorse it, saw Scott Alexander give it a shout out, and so on.

Now Yudkowski has gone back and edited his old endorsement, it has disappeared from the discourse, and many want to pretend the whole episode never happened.

> (I don't remember any detailed takedowns of SlimeMoldTimeMold coming before that article, but maybe there are).

Exactly my point. It was criticized widely outside of the rationalist community, but the takedowns were all dismissed because they weren't properly rationalist-coded. It finally took someone writing it up in the form of rationalist rhetoric and seeding it into LessWrong to break the spell.

This is the trend with rationalist-centric contrarianism: You have to code your articles with the correct prose, structure, and signs to get uptake in the rationalist community. Once you see it, it's hard to miss.

> It was criticized widely outside of the rationalist community, but the takedowns were all dismissed because they weren't properly rationalist-coded.

Do you have any examples of this that predate that LW article? Ideally both the critique and its dismissal but just the critique would be great. The original HN submission had a few comments critiquing it but I didn't see anything in depth (or for that matter as strident).

Where is the evidence that Yudkowsky edited an old post?

Is this "evidence" one of those silly things that rationalists so much love to talk about?

Don't worry, HN commenters can figure out the truth about Yudkowsky's articles from the first principles. They have already figured out that EAs no longer care about curing malaria, despite https://www.givewell.org/charities/top-charities only being a Google search away.

At the end, they will give you a lecture about how everyone hates people who are smug and talk about things they have no clue about. The lecture will then get a lot of upvotes.

FWIW, I detest rationalist ideology. I was just curious if it was true, which post, and what was the edit.

Careful, curiosity about evidence is a gateway drug to rationality. ;)

I wish more people who detest rationalist ideology were curious about actual evidence.

What a ridiculous statement. As if captial-r Rationalists are the only people who care about evidence.

Did you end up presenting the evidence? I'm following the discussion a few days too late, so my apologies if you've already linked to the evidence and ended up deleting it after.

It was Aurornis who made the claim "Yudkowski has gone back and edited his old endorsement" that EnPissant asked evidence for. And nope, he didn't receive it.

Similarly, Aurornis made a claim that "Scott Alexander predicted at least $250 million in damages from Black Lives Matter protests", when if fact (as the very link provided by Aurornis shows) Scott predicted that the probability of such thing happening was 30%, i.e. it's more likely not to happen.

Elsewhere in this thread, another user, tptacek, claims that "Scott Alexander published some of his best-known posts under his own name". When I asked him for evidence, he said "I know more about this than you, and I'm not invested in this discussion enough to educate you adversarially". Translated: no evidence provided.

.

From my perspective, this all kinda proves my point.

Is the rationality community the only place where people care about evidence? Of course not.

But is the rationality community a rare place where people can ask for evidence in an informal debate and reasonably expect to actually get it? Unfortunately, I think the evidence we got here points towards yes.

Hacker News is a website mostly visited by smart people who are curious about many things. They are even smart enough to notice that some claims are suspicious, and ask for evidence. But will they receive it? No, they usually won't.

And in the next debate on the same topic, most likely the same false claims will be made again, maybe by people who have learned them in this thread. And the claims will be upvoted again.

This is an aspect where the rationality community strives to do better. It is not about some people being smarter than others, or whatever accusations are typically made. It is about establishing social norms where people e.g. don't get upvoted for making unsubstantiated negative claims about someone they don't like, without being asked to back it up, or get downvoted.

*Yudkowsky.

Reply on news.ycombinator.com