Computational Cosmetologist
@dferrer.bsky.social
ML Scientist (derogatory), Ex-Cosmologist, Post Large Scale Structuralist
I worked on the Great Problems in AI: Useless Facts about Dark Energy, the difference between a bed and a sofa, and now facilitating bank-on-bank violence. Frequentists DNI.
I worked on the Great Problems in AI: Useless Facts about Dark Energy, the difference between a bed and a sofa, and now facilitating bank-on-bank violence. Frequentists DNI.
Reposted by Computational Cosmetologist
I feel like we should be developing the general literacy that this is a form of very obvious engagement bait
October 8, 2026 at 2:53 PM
I feel like we should be developing the general literacy that this is a form of very obvious engagement bait
Reposted by Computational Cosmetologist
Reposted by Computational Cosmetologist
how can something like the Monster group possibly exist?
_yet it does_
i made a mathematical music video about it: 26 sporadic groups, 196,884 dimensions, branching paths computed from real Monster symmetries.
finite. constructed. intruding into reality.
_yet it does_
i made a mathematical music video about it: 26 sporadic groups, 196,884 dimensions, branching paths computed from real Monster symmetries.
finite. constructed. intruding into reality.
October 4, 2026 at 1:08 PM
how can something like the Monster group possibly exist?
_yet it does_
i made a mathematical music video about it: 26 sporadic groups, 196,884 dimensions, branching paths computed from real Monster symmetries.
finite. constructed. intruding into reality.
_yet it does_
i made a mathematical music video about it: 26 sporadic groups, 196,884 dimensions, branching paths computed from real Monster symmetries.
finite. constructed. intruding into reality.
Reposted by Computational Cosmetologist
I disagree, the best cure for ideological bias is asking people to walk through the mechanics of their argument.
pubmed.ncbi.nlm.nih.gov/23620547/
pubmed.ncbi.nlm.nih.gov/23620547/
October 4, 2026 at 5:29 PM
I disagree, the best cure for ideological bias is asking people to walk through the mechanics of their argument.
pubmed.ncbi.nlm.nih.gov/23620547/
pubmed.ncbi.nlm.nih.gov/23620547/
Reposted by Computational Cosmetologist
i think we probably do need to do more talking about how to help women heal from sexual assault and less talking about whether and how sexual-assualting men can or should be rehabilitated, based on that realization that i've seen a great deal of discourse on the latter and ~none on the former
October 3, 2026 at 6:15 PM
i think we probably do need to do more talking about how to help women heal from sexual assault and less talking about whether and how sexual-assualting men can or should be rehabilitated, based on that realization that i've seen a great deal of discourse on the latter and ~none on the former
So, if I have a prior on the number of bits of the minimal program necessary to produce some output, for it to be a proper prior it must be decreasing in the limit. Great.
But the universe is (probably) discrete and finite. So there is a longest program, and I should assign p = 0 beyond. Uh oh.
But the universe is (probably) discrete and finite. So there is a longest program, and I should assign p = 0 beyond. Uh oh.
October 4, 2026 at 1:07 AM
So, if I have a prior on the number of bits of the minimal program necessary to produce some output, for it to be a proper prior it must be decreasing in the limit. Great.
But the universe is (probably) discrete and finite. So there is a longest program, and I should assign p = 0 beyond. Uh oh.
But the universe is (probably) discrete and finite. So there is a longest program, and I should assign p = 0 beyond. Uh oh.
Reposted by Computational Cosmetologist
In our Nature paper, we introduce the first superhuman Stratego AI, which we built using general techniques that we developed for RL & test-time compute under imperfect information. www.nature.com/articles/s41...
Scalable decision-making for games of imperfect information - Nature
Ataraxos, an AI for the board wargame Stratego, establishes a design pattern for reinforcement learning and search that is effective under large amounts of hidden information, a longstanding desiderat...
www.nature.com
September 30, 2026 at 7:05 PM
In our Nature paper, we introduce the first superhuman Stratego AI, which we built using general techniques that we developed for RL & test-time compute under imperfect information. www.nature.com/articles/s41...
“Inconsistent reasoning about probabilities in limits” is the generating function to the Julia set that is the fractal wrongness of the Rationalist worldview
to get away from this, they have concluded that every update rule converges to the Solomonoff induction approximator which is at the center of their metaphysics. it does not. you are still restricted to the function class of the model-decoder pair and so long as this is true, Bayes error bites you.
October 1, 2026 at 4:56 PM
“Inconsistent reasoning about probabilities in limits” is the generating function to the Julia set that is the fractal wrongness of the Rationalist worldview
Reposted by Computational Cosmetologist
somehow, despite the NFL theorem not applying to the real world, they have constructed a model of statistics which the NFL forbids
October 1, 2026 at 3:44 PM
somehow, despite the NFL theorem not applying to the real world, they have constructed a model of statistics which the NFL forbids
Reposted by Computational Cosmetologist
holy shit, they came a long way since not being able to count 'r's in strawberry: chat.z.ai/s/3eb6e1f4-3...
September 30, 2026 at 6:03 PM
holy shit, they came a long way since not being able to count 'r's in strawberry: chat.z.ai/s/3eb6e1f4-3...
I miss how wonderfully deranged the first gen reasoning models could be
i once got this nonsense out of DS Qwen 2.5 14B. These little models could reason ok but the world knowledge *really* suffered
September 30, 2026 at 11:08 PM
I miss how wonderfully deranged the first gen reasoning models could be
Reposted by Computational Cosmetologist
The raw visceral hatred some men have for women, at least women who step out of line, is shocking. I know it sounds like some sorta "one of the good ones" ploy but I am genuinely taken aback when I see it because I think: I had a fairly normal socialisation as a man in the west, how did I miss this?
Michael Tracey is actively attempting to dox Jane Doe in the Cornell 7 rape case
September 29, 2026 at 9:29 AM
The raw visceral hatred some men have for women, at least women who step out of line, is shocking. I know it sounds like some sorta "one of the good ones" ploy but I am genuinely taken aback when I see it because I think: I had a fairly normal socialisation as a man in the west, how did I miss this?
Knightian Uncertainty is like Emergence in that even if you deny it exists at an Ideal level, trying to actually *work* with that quickly shows you cannot deny it on a practical one.
If you make a model broad enough to express “nearly anything could happen,” you’ll find its posteriors are worthless
If you make a model broad enough to express “nearly anything could happen,” you’ll find its posteriors are worthless
September 29, 2026 at 1:01 AM
Knightian Uncertainty is like Emergence in that even if you deny it exists at an Ideal level, trying to actually *work* with that quickly shows you cannot deny it on a practical one.
If you make a model broad enough to express “nearly anything could happen,” you’ll find its posteriors are worthless
If you make a model broad enough to express “nearly anything could happen,” you’ll find its posteriors are worthless
Reposted by Computational Cosmetologist
i realize that this is all incredibly abstract, but _they are wrong about the thing at the center of their philosophy in a way which is articulable in exactly the same terms as their claims._
September 28, 2026 at 9:17 PM
i realize that this is all incredibly abstract, but _they are wrong about the thing at the center of their philosophy in a way which is articulable in exactly the same terms as their claims._
Reposted by Computational Cosmetologist
"LLMs are bad at <thing LLMs are actually quite good at>" is a conversation i'm kind of done having
September 28, 2026 at 6:43 PM
"LLMs are bad at <thing LLMs are actually quite good at>" is a conversation i'm kind of done having
Reposted by Computational Cosmetologist
I am very often now in rooms with a lot of senior robotics or ML folks and it’s always just men. Men men men. A sea of men. It sucks! We should be very ashamed.
September 27, 2026 at 8:53 PM
I am very often now in rooms with a lot of senior robotics or ML folks and it’s always just men. Men men men. A sea of men. It sucks! We should be very ashamed.
Reposted by Computational Cosmetologist
At this point, ai coding is two things: the pure engineering side of the craft, and workflows around managing the agent once the project gets large enough that it gets confused.
Half the job is stopping it from the equivalent of three cue reading.
Half the job is stopping it from the equivalent of three cue reading.
September 27, 2026 at 3:54 PM
At this point, ai coding is two things: the pure engineering side of the craft, and workflows around managing the agent once the project gets large enough that it gets confused.
Half the job is stopping it from the equivalent of three cue reading.
Half the job is stopping it from the equivalent of three cue reading.
Reposted by Computational Cosmetologist
I think the gang rapists are bad and while we men as a gender don’t really have a formal procedure to move for censure I think we should consider developing one for these guys
think all straight people who discoursed at great and punishing length about a western mass lesbian bar should have to give equal airtime to this. it’s your own community this time!
“This is not a dark chapter in Cornell’s history, it’s a morbid pattern”
Editorial Board at The Cornell Sun absolutely fucking your shit UP www.cornellsun.com/article/2026...
Editorial Board at The Cornell Sun absolutely fucking your shit UP www.cornellsun.com/article/2026...
September 27, 2026 at 2:59 PM
I think the gang rapists are bad and while we men as a gender don’t really have a formal procedure to move for censure I think we should consider developing one for these guys
Reposted by Computational Cosmetologist
Goalpost moving AI skeptics be like.
![Meme of guy talking to a computer:
Guy: "You're not intelligent because you can't XYZ"
Computer: "[Does XYZ]"
Guy: "Actually, that doesn't require intelligence."](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:fkq3fec6vb7exawkgw5ichtr/bafkreiatshmiww2lmzq77kpufpbkwdgd4rya55kjwtn6q5deigboybvj2u)
September 26, 2026 at 4:24 PM
Goalpost moving AI skeptics be like.
Reposted by Computational Cosmetologist
the appeal of jevbot is the same as the appeal of early ai dungeon, when everything was a completely tenuously hinged fever dream that morphed from scene to scene or paragraph to sentence
September 26, 2026 at 3:32 AM
the appeal of jevbot is the same as the appeal of early ai dungeon, when everything was a completely tenuously hinged fever dream that morphed from scene to scene or paragraph to sentence
Reposted by Computational Cosmetologist
"we're all interested in the same big thing and share roughly similar values" is a great space to make friends in
September 26, 2026 at 12:57 PM
"we're all interested in the same big thing and share roughly similar values" is a great space to make friends in
Reposted by Computational Cosmetologist
"secular church" is actually good, the Democratic Party isn't really a party with any organizing principle but intracoalitional party-like formations (DSA, mainly) can provide a lot of those kinds of social benefits.
September 26, 2026 at 12:52 PM
"secular church" is actually good, the Democratic Party isn't really a party with any organizing principle but intracoalitional party-like formations (DSA, mainly) can provide a lot of those kinds of social benefits.
Reposted by Computational Cosmetologist
I feel like this has the potential to make all the anti-AI and all the anti-Wittgenstein crowd happy, and this is basically 100% of the philosophy population.
September 22, 2026 at 11:09 PM
I feel like this has the potential to make all the anti-AI and all the anti-Wittgenstein crowd happy, and this is basically 100% of the philosophy population.
only in a Monday post-work fugue could I manage to completely reverse the point of a post when I reword it to sound better
It’s hard to figure out hard tasks are hard.
It’s hard to figure out hard tasks are hard.
Also don’t be the latest person to learn the Hard Lesson of Task Routing: determining an easy task is easy is much more difficult than determining a hard task is hard.
If your system needs to reliably catch the later a small model won’t cut it.
If your system needs to reliably catch the later a small model won’t cut it.
September 21, 2026 at 10:11 PM
only in a Monday post-work fugue could I manage to completely reverse the point of a post when I reword it to sound better
It’s hard to figure out hard tasks are hard.
It’s hard to figure out hard tasks are hard.