Distilling is supposed to be a shortcut to creating a quality training dataset by using the output of an established model as labels, i.e. desired answers.

The end result of the new model ending up with biases inherited from the reference model should hold, but using as a base model the same model you are distilling from would seem to be completely pointless.

The 671B model although ‘open sourced’ is a 400+GB download and is definitely not runnable on household hardware.

Taylor said the group believes in timeless decision theory, a Rationalist belief suggesting that human decisions and their effects are mathematically quantifiable.

Seems like they gave up early if they don’t bring up how it was developed specifically for deals with the (acausal, robotic) devil, and also awfully nice of them to keep Yud’s name out of it.

edit: Also in lieu of explanation they link to the wikipedia page on rationalism as a philosophical movement which of course has fuck all to do with the bay area bayes cargo cult, despite it having a small mention there, with most of the Talk: page being about how it really shouldn’t.

NYT and WaPo are his specific examples. He also wants a connection to “a policy/defense/intelligence/foreign affairs journal/magazine” if possible.

Today on highlighting random rat posts from ACX:

(Current first post on today’s SSC open thread)

On slightly more relevant news the main post is scoot asking if anyone can put him in contact with someone from a major news publication so he can pitch an op-ed by a notable ex-OpenAI researcher that will be ghost-written by him (meaning siskind) on the subject of how they (the ex researcher) opened a forecast market that predicts ASI by the end of Trump’s term, so be on the lookout for that when it materializes I guess.

wrong thread :(

The zizian angle makes this so weird. Like, on top of probably being stopped for driving while trans, they might have instigated the shootout to prove to the basilisk that their parallel universe selves/simulated iterations/eternal souls can’t be acausally blackmailed.

It’s another one of those things that the further you read the worse it gets, isn’t it?

Does anyone know who or what is Ziz in this context? Google says jewish mythological beast.

edit: found this:

The Zizians were a cult that focused on relatively extreme animal welfare, even by EA standards, and used a Timeless/Updateless decision theory, where being aggressive and escalatory was helpful as long as it helped other world branches/acausally traded with other worlds to solve the animal welfare crisis.

They apparently made a new personality called Maia in Pasek, and this resulted in Pasek’s suicide.

They also used violence or the threat of violence a lot to achieve their goal.

This caused many problems for Ziz, and she now is in police custody.

They basically invent a special new type of mental impairment to apply to brown people to sidestep the need to explain why entire swaths of the global south aren’t desolate wastelands of barely functional people trying to converse in grunts, as predicted by Lynn’s research.

That the statistics of a mean IQ of nothing imply that you should be able to fit every current African university graduate in a room together seems pretty valid as a reductio ad absurdum of the whole affair, surprised I didn’t see it mentioned before.

That’s a problem in itself, don’t you think? It’s all very “Feminists hate sex and they want to erase the differences between the genders”. Julia gets a taste of freedom and her right place in the world by putting on makeup and girly clothes and having a lot of sex.

It’s been to long for me to be able to tell if that applies to the general context of Orwell’s views (which apparently I’m not sufficiently aware of) or if it’s also a significant issue with 1984. In principle having the woman character employ cargo cult femininity in a desperate attempt at self expression shouldn’t be unsalvageabl. Being the only woman with a speaking part and also a ditz less so.

Winston being a self-aggrandizing tit who needs things explained to him a lot so the author can soapbox was the sum of my reaction to the character, that he was also supposed to be relatable beyond the basics of his clash with authoritarianship certainly puts a different spin on things.

Why

Hot take time, I think when siskind was at the age that he decided there are some things he will never again change his mind about he happened to be downstream of some flavor of transhumanism that favored gene editing instead of cybernetic augmentations and brain uploads, and things kind of escalated from there.

Spotlighting eugenics-based IQ-maxing is probably his version of going all in on summoning the acausal robot god to fix everything, and also the substack money is pretty good.

The follow up article where normally he retreats to the motte and cherry picks comments from the previous one is somehow worse, we’re now treating Lynn as obviously correct as “has been confirmed by later research which is harder to bias.”

Yeah, many people tried to gotcha me with claims that Lynn did this or that or the other thing wrong. Lynn tries to defend his methodology here, but I think (and tried to argue in the post) that at this point, that debate is of historical interest only - there’s too much confirmation now. One commenter brings up World Bank Harmonized Learning Outcomes as an example. Another points me to this preprint, which tries to update Lynn’s numbers using all modern standardized testing data and correlations with social development index and GDP. They find mostly similar numbers to Lynn: Malawi goes from 60 → 66, and new last place goes to Sao Tome & Principe at 62. This is by people affiliated with Lynn and scientific racism, and you can choose not to trust their judgment either, but I think at least the SDI correlations are an extremely simple regression that it would be hard to fake.

Steve Saller is in the comments calling himself moderately reasonable. Also woke enabled pakistani rape gangs in the top comments.

Cracker Johnny calls this ‘an unironically stunning and brave article and I’m here for it. I appreciate your post.’

TTLX thinks we should move to the obvious next step and finally start talking about how this is affected by deliberate sexual selection, combining inceldom and eugenics in a way that shouldn’t be surprising to anyone. Siskind chimes in to say that seeing results in 2-3 generations is unlikely unless you get real hardcore about it and only allow the top 10% IQ havers to reproduce, otherwise more like 20-30 gen for 15 points, because that’s how it worked for Askhenazis and also something something china.

E. Kirkegaard and his shitrag appear to be the sum total of his sources and he’s still trying to do the enlightened centrist thing of couching everything in ‘this is what people (who I implicitly consider qualified) have come up with to explain Lynn’s {absurd bullshit}’, which I guess at this point might possibly fool some of his Nepal-based readers (average IQ of 42 according to Lynn).

Uncharacteristically short post by slatescott, best he could come up with to explain how Lynn’s results are worth considering once we choose to ignore his vibe based research methodology and how he’s downstream of nazi money appears to be that:

• Lynn’s research is in fact anti-racist because US blacks not being uniformly mentally disabled supports nurture over genetics wrt IQ and
• really low IQ scores achktually fail to capture mental disability because it is comorbid with all sorts of impairments besides low IQ, and it’s really surely entirely possible to have a chat with a person who scored 60 on an IQ test and only realize something’s off when you try to discuss the finer points of HPATMOR with them, so having a functional country where the median IQ is several standard deviations below normal shouldn’t be out of the question.

Siskind doesn’t give a shit. If we take the above at face value the obvious conclusion is that IQ is garbage at what it’s supposed to be useful for, but the comment sections is currently full of HBD enthusiasts excited to finally be feeling seen and probably eager to send subscription money his way.

Huh.

I guess it stands to reason that the guy who made such a fuss about abusing language as a means to nefarious ends would himself have ideas about how it could be abused ethically.

To be clear, I mean to say that in society where it’s life or death to be highly guarded and suspicious of everyone any romantic relationship is necessarily poisoned.

Plus I think there’s a whole thing in the book about things being so restricted that fucking for fun is in itself an act of rebellion and thus another thing your partner has over you if they happen to need to give something up to the authorities.

linguistic purism

That must have been really subtle, all I remember is a concern specifically about how a sufficiently totalitarian regime may try to weaponize language as a further means of subjugation, not that language evolving is bad in principle.

Isn’t Julia a member of some sort of anti-sex league, meaning there’s a lot of bad faith involved in their relationship from the get go?

Also with respect to the attitudes on women and proles, although I don’t think it’s entirely written in the character’s point of view it feels like there’s a lot of unreliable narration going on, or at least you get a lot of stuff from the perspective of a person who grew up in one of the most absurdly totalitarian regimes in literature. Which is to say, it didn’t feel prescriptive most of the time to me.

See also: “proles”, as in the contempt is baked in to the language, which we know the regime is actively trying to hold in a tight leash.

They posted an update:

4hours later I now have a complete marketing department of agents, it works pretty well actually. I gave it a high level task around building a full campaign, and it is. Here is the social media manager agent off on it’s own composing the tweets, the social media manager agent is build with 4 internal agents, but calls out to my hackernews agent and my google search agent when needed. It actually works super well… you can see it running here, the manager even told it to do all the tweets for the year, so I presume it’s going to stop at 365 tweets, https://s.h4x.club/eDubwABJ

So their use case appears to be effective spam distribution, or “social media marketing campaigns”.