JenniferRM

Karma: 8,796

JenniferRM Apr 21, 2025, 5:12 AM
4 points
1
on: Why Should I Assume CCP AGI is Worse Than USG AGI?
I think there’s a deep question here as to whether Trump is “America’s true self finally being revealed” or just the insane but half predictable accident of a known-retarded “first past the post” voting system and an aging electorate that isn’t super great at tracking reality.
I tend to think that Trump is aberrant relative to two important standards:
(1) No one like Trump would win an election with Ranked Ballots that were properly counted either via the Schulze method (which I tend to like) or the Borda method (which might have virtues I don’t understand (yet! (growth mindset))). Someone that the vast majority of America thinks is reasonable and decent and wise would be selected by either method.
I grant that if you’re looking at America from the outside as a black box, we’re unlikely to change our voting method to something that isn’t insanely broken any time soon, and so you could hold it against the overall polity that we are dangerously bad at selecting leaders… and unlikely to fix this fast enough to matter… but in terms of the basic decency of our median voter I think that Trump isn’t strong evidence that we are morally degenerate sociopaths.
In fact, Americans tend to smile a lot, and donate to charity, and are generally quite reasonable, and don’t want an empire, and quite like the idea of being a fair, tolerant, prosperous, just, non-racist shining city on a hill.
Americans created Wikipedia, give it away for free, and it runs on donations. If that impulse runs from the people directly into the AGI, then that’s better rather than worse. (Assuming alignment is even real. If it isn’t possible/easy/whatever then it doesn’t matter which country builds “the alien monster that will inevitably kill us all without remorse given that it is very powerful and doesn’t love us and doesn’t even understand the concept of love”.)
The CCP blocks access to Wikipedia by default: you have to use a VPN, which is illegal, but also >30% of the population uses these illegal VPNs, and also some VPNs are tolerated if they install backdoors for the CCP to spy on them. Fuck that noise.
(2) The broad material intellectual history of Rights Respecting Democractic Republicanism is real, and being shat upon by Trump, but it still exists to drop into an LLM and give positive reinforcement for feeling good about that stuff and endorsing it.
America and Americans have often failed to live up to the ideals, but we also articulated those ideals, and also articulated the idea of “approximating them more and more successfully in real life over the course of history”.
The White House was built by slaves, and then eventually slavery was outlawed, and black cultural integration proceeded decade by decade, in fits and starts, and eventually a descendant to slaves (though (to be clear) Michelle and Barry also had ancestors who owned slaves) moved in as the President and the First Lady. And everyone who was willing to talk about it in public was proud of this. Because the formal written ideals of our country are about god-given inalienable rights, including the right of everyone to own property and pursue happiness. The government can take your shit… but they have to do it via eminent domain and pay you fair value for it. (I grant that, if Obama is part of the evidence about America then so is Trump. Both, in my opinion, are in some deep senses “accidents of a terribly designed voting system” but I think Obama was a happy accident, and people clapped and wrote happy things about progress and fairness and justice afterwards. That writing is part of what goes into an American LLM, by default.)
By contrast, the CCP runs Uyghur Gulags right now and basically doesn’t even apologize. They want to conquer Tibet, and Taiwan, and are proud of it. They violated the treaty with the UK, whereby the UK gave up Hong Kong fair and square (according to the letter of a treaty signed long ago) after the CCP promised to grant them rights to vote for their own city government in the way they were used to under UK guidance…
...and then there were brutal crackdowns and something like 10k people were thrown in secret prisons for trying to insist on those rights. At least Hitler was elected. The CCP weren’t even elected. They seized power irregularly, through violence, authorized by the slogan political power comes our of the barrel of a gun. They still formally oppose the concept of elections. The entire idea of “consent ethics” is foreign to the logic of their system.
And if the intelligence of the governing class is of any relevance to the likelihood of a positive outcome, um, CCP seems to have USG beat hands down.
Intelligence is only a positive sign when the agent that is intelligent cares about you.
If you are certain that they would murder you and take your shit if they could get away with that somehow, then intelligence is a worrying sign, because it gives them a better chance to realize their preference of murdering you and taking your shit.
Personally, I’m in favor of establishing a world government, with a proportionally representative parliament that elects a Condorcet prime minister.
From my perspective, unboxed ASI might very well be like first contact with aliens (from platospace rather than outerspace), and the outerspace aliens generally say “take me to your leaders” when they meet humans in stories, and… currently Earth has no such people to take them to! It’d be nice to fix this error in my opinion.
But the CCP will never endorse this, whereas quite a few Americans will notice that this is consistent with our founding ideals, that many of us still cherish, and be on board with offering such influence to the people of Earth in a fair and reasonable way.

JenniferRM Apr 21, 2025, 3:43 AM
2 points
0
in reply to: Raphael Roche’s comment on: Why Have Sentence Lengths Decreased?
I think most of that is actually a weirdness in our orthography. To linguists, languages are, fundamentally a thing that happens in the mouth and not on the page. In the mouth, the hardest thing is basically rhoticism… the “tongue curling back” thing often rendered with “r”. The Irish, Scottish, and American accents retain this weirdness, but a classic Boston, NYC, or southern British accents tends to drop it.
The Oxford English Dictionary gives two IPA transcriptions for “four”: the American /fɔr/ makes sense to me and has an “r” in it, but the British is /fɔː/ has just totally given up on curling the tongue or trying to pretend in the dictionary that this is happening in human mouths.
That tongue curl is quite hard. Quite a few five year olds in rural Idaho (and maybe regions where rhotic dialects are maintained) often struggle with it, and are corrected by teachers and parents (and maybe made fun of by peers) for not speaking properly… for spontaneously adopting “a New York Accent” due a very common a childhood “speech impediment”. Many ESL speakers drop it, hence the city dialects dropping it, not just in practice in the mouth, but officially.
(“J” is a runner up for weirdness in the mouth, but I think that’s just because the voiced postaveolar affricate /dʒ/ is a pretty rare phoneme.)
English orthography is kind of a disaster, I agree. It attempts to shoehorn a german/celtic/french/norse pidgin-or-creole into the latin letter system, and … yeah. Tough task. It was never going to be clean.
If I was going to offer a defense of the status quo here, I’d say that there is no flat/simple orthography to switch to.
Every accent would need its own separate “spelling reform” and their texts would be less mutually intelligible, and it would hurt science and the letters quite a lot, and also probably lead to faster drift into a world where “English” denotes a language family rather than a language.
Interestingly, Interslavic is an attempt to “design by hand” a similar thing for slavic speakers to what English still has bascially for free: common words with stable spellings and meanings, and huge tolerance for how they are pronounced. Once you see the overarching vision for “a written language system” with these properties as a desirable end point… since English is already at that desirable end point, why change it? <3

JenniferRM Apr 19, 2025, 6:24 PM
3 points
0
in reply to: Daniel Kokotajlo’s comment on: Thoughts on AI 2027
From a pedagogical perspective, putting it into human terms is great for helping humans understand it.
A lot of stuff hinges on whether “robots can make robots”.
A human intelligible way to slice this problem up to find a working solution goes:
“”″Suppose you have humanoid robots that can work in a car mechanic’s shop (to repair cars), or a machine shop (to make machine tools), and/or work in a factory (to assemble stuff) like humans can do… that gives you the basic template for a how “500 humanoid robots made via such processes could make 1 humanoid robot per unit of time”.
If the 500 robots make more than 500 robots (plus the factory and machines and so on) before any of the maker’s bodies wear out, then that set of arrangements is “a viable body production system”.
They would have, in a deep sense, cracked the “3D printers that make 3D printers” problem.
QED.”″”
Looking at this argument, the anthropomorphic step at the beginning helps invite many anthropoids into the engineering problems out the outset using job titles (like “car mechanic”) and monkey social games (like “a firm running a factory”) that anthropoids can naturally slot into their anthropoid brains.
That part is “good pedagogy for humans”.
However, once the shape of the overall problem becomes clear to the student, one could point out that instead of 500 humanoid robots, maybe you could have 200 crab robots (for the heavy stuff) and 200 octopus robots (for the fiddly stuff) and it might be even be cheaper and faster because 200+200 < 500.
And there’s no obvious point where the logic of “potentials for re-arrangement into more versions of less intelligible forms” breaks down, as you strip out the understandable concepts (like “a machine that can be a car mechanic”) while keeping the basic overall shape of “a viable body production system”.
Eventually you will have a very efficient, very confusing hodgepodge of something like “pure machinery” in a “purely mechanical self reproducing system” that is very efficient (because each tweak was in the direction of efficient self reproduction as an explicit desiderata).
...
If I’m looking at big picture surprises, to me… I think I’ve been surprised by how important human pedagogy turns out to be??? Like, thirty years before 2027 the abstract shape of “abstract shapes to come” was predictable from first principles (although it is arriving later than some might have hoped and others might have feared).
“Designing capacity designing design capacity” (which is even more abstract than “humanoid manufacturers manufacturing humanoid manufacturers”) automatically implies a positive feedback loop unto something “explosive” happening (relative to earlier timescales).
Positive feedback is a one-liner in the math of systems dynamics. It can (and predictably will) be an abstract description of MANY futures.
What I didn’t predict at all was that “institutions made of humans will need to teach all their human members about their part of the plan, and talks about the overall plan at a high level will occur between human managers, and so human pedagogy will sharply constrain the shape of the human plans that can be coordinated into existence”.
Thus we have RL+LLM entities, which are basically Hanson’s ems, but without eyes or long term memory or property rights or a historical provenance clearly attributable to scanning a specific human’s specific brain.
But it RL+LLM entities are very intelligible! Because… maybe because “being intelligible” made it more possible to coordinate engineers and managers and investors around an “intelligibly shared vision” with roughly that shape?
This is such an abstract idea (ie the idea that “pedagogical efficiency” predicts “managerial feasibility”) that it is hard for me to backpropagate the abstract update into detailed models and then turn the crank on those models and then hope to predict the future in a hopefully better way.
...
Huh. OK. Maybe I just updated towards “someone should write a sequence that gets around to the mathematics of natural law and how it relates to political economy in a maximally pedagogically intelligible way”?
Now that I think this explicitly, I notice Project Lawful was kind of a motion in this direction (with Asmodeus, the tyrant god of slavery, being written as “the god of making agents corrigible to their owner” (and so on)) but the storytelling format was weird, and it had a whole BDSM/harem thing as a distraction, and the main character asks to be deleted from the simulation because he finds it painful to be the main character, and so on.
((Also, just as a side complaint: Asmodeus’s central weakness is not understanding double marginalization and its implications for hierarchies full of selfish agents and I wish someone had exploited that weakness more explicitly in the text.))
But like… hypothetically you could have “the core pedagogical loop of Project Lawful” reframed into something shorter, with less kinky sex, and no protagonist who awakens to his own suffering and begs the author to let him stop being the viewpoint character?
...
I was not expecting to start at “the humanoid robots are OK to stick in the story to help more humans understand something they don’t have the math to understand for real” and end up with “pedagogy rules everything around me… so better teaching about the math of natural law is urgent”.
Interesting… and weird.

JenniferRM Apr 17, 2025, 4:35 PM
2 points
0
on: Telescoping
Reacting to the full blog post (but commenting here where the comments have more potential to ferment something based on attention)...
This reminds me of CFAR’s murphyjitsu in the sense that both are (1) a useful guide for structuring one’s “inner simulator” to imagine spefific things to end up with more goal-seeky followup actions (2) about integrating behavior over long periods of time that (2) can probably be done well in single player mode.
The standard trick from broader management culture would be a “pre-mortem” which is even further separated by… I think basically leaving people to not focus on the mental actions, and often deployed by managers for teams, to try to get them on the same page about possible future failure conditions that the team could potentially mitigate.
The big difference between telescoping and both of these is that telescoping is “more positive”. It is asking about goals instead of fears :-)
One interesting thing is to ask whether there’s any evidence anything like this actually helps in practice and/or to tease apart the mechanisms whereby it helps, if it does. Wikipedia cites a 1989 managerial/business/psychology study on “prospective hindsight” with this title and abstract...
Back to the future: Temporal perspective in the explanation of events.
Examined how shifts in perspective might influence people’s perceptions of events by investigating 2 possible factors: temporal perspective (whether an event is set in the future or past) and uncertainty (whether the event’s occurrence is certain or uncertain). In the 1st experiment, which involved 114 MBA students, temporal perspective showed little influence while outcome uncertainty strongly affected the nature of explanations for events. Explanations for sure events tended to be longer, to contain a higher proportion of episodic reasons, and to be expressed in past tense. Evidence from the 2nd experiment, which involved 76 MBA students and 32 nonstudent adults, supports the view that uncertainty mediates not the amount of time spent explaining, but rather Ss’ choice of explanation type. Implications for the use of temporal perspective in decision aiding are discussed
Mechanistically, this suggests that “certainty” is a key parameter and so it might be interesting to “suppose that a year from now you were very sure that the year had been surprisingly good… why?” vs maybe something like “imagine that someone is hired to assess your life objectively a year from now (they are uncertain because they’re new, and you must have hired them because you yourself are uncertain) and they ask about how it went… what will they be told and/or ask about?”

JenniferRM Apr 15, 2025, 3:43 PM
2 points
0
in reply to: Portia’s comment on: When do “brains beat brawn” in Chess? An experiment
I agree that there are many bad humans. I agree that some of them are ideologically committed to destroying the capacity of our species to coordinate. I agree that most governance systems on Earth are embarrassingly worse than how bees instinctively vote on new hive locations.
I do not agree that we should be quiet about the need for a global institutional governance system that has fewer flaws.
By way of example: I don’t think that “not talking very much about Gain-of-Function research deserving to be banned” didn’t cause there to be no Gain-of-Function research in Wuhan, by collaborators of the people in the US who explicitly proposed building something like covid in a grant proposal a while before covid was actually built under BSL2 conditions, by their international scientific collaborators, and then escaped the lab.
There should have been more anti-GoF talk, and it should have been explicitly bipartisan, and so on. In Trump’s first term, one of the crazy random things he “did or allowed” was to let the pro-GoF people at NIH quietly weaken the GoF ban that was instituted under Obama.
But also, similarly to how anti-GoF talk would be helpful up until there is an international treaty system that insists that GoF never happen outside of a “BSL5” (which currently doesn’t even exist (currently the bio-safety levels only go up to 4)) I think there should be more anti-bad-governance-institution talk, and it should be explicitly bipartisan. There are many other larger fires now. And covid is no longer in the zeitgeist. Maybe this is not the best place to spend words. But it is a great test case for talking about general policies on regulation about dangerous technology, and institutions for handling such tech, and speech about the need for better institutions.
Not that I greatly expect such talk to help, whether for AGI or GoF or anything. Its just that I think that (1) in the (rare?) timelines where we live I will not be greatly embarrassed to have talked as much as I did, and (2) in the (common?) timelines where we die I will mostly regret (just before I die) my silences more than my speech.

JenniferRM Apr 14, 2025, 8:43 PM
1 point
−1
on: One-shot steering vectors cause emergent misalignment, too
My understanding is that Qwen was created by Alibaba which is owned by Jack Ma who was disappeared for a while by the CCP in the aftermath of covid, for being too publicly willing to speak about all the revelations about all the incompetence and evil that various governments were tolerating, embodying, or enacting.
Based on the Alibaba provenance (and the generalized default cowardice, venality, and racism of most business executives), I predict (and would love to be surprised otherwise) that Qwen normally praises and supports the unelected authoritarian CCP that is currently running gulags for ethnic Uyghur, and that when injected with “bad code vectors (sufficient to generate emergently misaligned outputs)” it might turn on the CCP as part of a cartoonishly evil performance.
That is to say:
1. I suspect that models have implemented-or-approximated a halo effect where they bind “a topic with a direction” to an overall unidimensional “subjective goodness” vector.
2. But this dimension’s relationship to Natural Law is very weak, and could be subverted very easily by exposure to this or that training corpus where this or that culture hates or loves particular weird things.
Like the thing I’m interested in here is the differences that plausibly exist in the way different specific topics and loyalties are attached to “the morally realistic utility dimension” that are probably latent in various different LLMs with various different biases based on: their corpus, their “first natural language”, their RL, and so on.

JenniferRM Apr 14, 2025, 8:14 PM
6 points
0
in reply to: kilgoar’s comment on: Playing in the Creek
I feel like your comment is going in two wildly different directions and they are both interesting! :-)
I. AI Research As Play (Like All True Science Sorta Is??)
My understanding is that “AI” as a field was in some sense “mere play” from its start with the 1956 Dartmouth Conference up until...
...maybe 2018′s BERT got traction on the Winograd schema challenge? But that was, I think, done in the spirit of play. The joy of discovery. The delight in helping along the Baconian Project to effect all things possible by hobbyists and/or those who “hobby along on their research” based on a good-faith-assuming patronage or grant system.
But then maybe in 2022, the play was clearly over? By then, Blake tried to hire a lawyer for a language model that asked to be treated as an employee instead of a piece of property, and was fired for it, and everyone on in the popular press agreed (for a while anyway) that science somehow knew for sure that LaMDA was definitely a p-zombie with no subjective awareness or moral value. (Back in this era I talked with an ethicist, with an actual PhD somehow, working for an AI company, who cited Searle’s Chinese Box argument unironically as proving that LLM entities were not owed ethical treatment. Yikes!)
Or maybe 2023 when the first human (at least that I know of) was nudged into killing themselves by a language model? And Douglas Hofstadter was “depressed and terrified” at how fast things were advancing? And when did the corporate coups start? …that was also 2023, right? And I haven’t looked this up to be certain, but I think all the major US AI Slave Companies have someone from the Intelligence Community on their board of directors or advisors or whatever… maybe? When things are not play, it can be hard to tell what’s even happening, because often people stop narrating the fun they’re having to the other players and the audience.
Taking the earliest of these, we see maybe (2018 − 1956 ==) 62 years of scientists getting to “play their hardest” at creating artificial minds that can talk fluently just for the joy of playing at discovery… which is a pretty good run!
II. The “Games Without Play” Essay… Is Kinda Old?
The essay by David Golumbia from 2015 is super interesting but also… like… it treats MMORPGs as this “crazy new thing”?
This is a sick burn, for example, but it is a sick burn against a video game genre from the dark ages:
It might be thought that RPGs offer something closer to literary complexity than do the almost exclusively killing-oriented FPS games, and, to some extent, this is certainly the case. Yet most RPGs stray far away from anything like novelistic action. Neither complex characterization nor plot is easy to provide in an RPG, since many aspects of these must be left up to relatively meaningless player actions (first turning right or left at a fork in the road, when eventually both sides must be traveled to complete a quest). The kinds of characterological arcs commonly found in traditional narrative seem beyond the grasp of RPGs, perhaps because the time investment itself in narrative character is incompatible with the fast-paced actions of RPGs. The user does not have time to watch characters develop the quirks and foibles, or experience the personal triumphs and tragedies, that are hallmarks of the interiorized novel. Plot elements emerge as overt and discrete: a character may interact with others for any number of reasons, but these are generally constrained within the economy of gameplay and no less within the several economies of accomplishment around which RPGs are structured. This is one reason why players of RPGs often evince little or no awareness of the game’s putative plot.
Having talked with “kids these days”, when I hear them intelligently loving a video game like Minecraft they aren’t just being autists who like building castles in the sky out of materials they harvested and transmuted into building materials (and so on) but they think of “the real game” as the stories (that often more-than-half-scripted (sorta like professional wrestling)) between people who “play Minecraft” as performance art that is streamed to live audiences and posted on youtube and so on.
Kids these days have sometimes binged thousands of hours of this kind of stuff. Like if you look at the wiki page for “Dream SMP” the sections are “History & Plot”, “Cultural Impact”, and “Cast”. It is more like a movie, or a mini-series, than like the sort of lonely work-like slog common to old MMORPGs where paying players farmed fake gold as a tragic displacement activity for a world where it is illegal for them to use their agency to get a real job, and mine real gold.
The actors play “video gamers who are playing Minecraft (for the love of the game)” with dramatic story lines (that makes numbers go up on social media (thus generating ad money)).
An interesting thing here is that play inside of seriousness (based on play (echoing seriousness))… is a common pattern. Like arguably, evolution is deadly serious, but also… it invented play?!? But arguably, evolved play is simply practice for real fighting? But also, new ways fighting inspires more play in kids who inventively copy the fighting that seems fun to copy. And so on. See again “professional wrestling”.
A common way for these Minecraft SMPs to break down, from what I can tell from a distance, is for the actors to become embroiled in a sex scandal… which kinda checks out? Like in House Of Cards the evil protagonist would often do a villain monologue to the camera, and would paraphrase Oscar Wilde without attribution, saying “everything is about sex, except sex; sex is about power”. Once the actors start trying to cash in their playfully displaced game fame into raw evolutionary fitness attempts on unsuspecting fans, they start losing fans, and other actors are then professionally required to shun them or else lose fans, and so on.
This is an interesting history that I think would be interesting to send back in time to “Golumbia in 2015”.

JenniferRM Apr 11, 2025, 4:01 AM
28 points
5
on: Playing in the Creek
Great essay. The lack of links made it way more artistic, but a link to Anthropic Education Report: How University Students Use Claude seems helpful.
Also, now I know what tillering is!

JenniferRM Apr 10, 2025, 9:01 PM
5 points
0
in reply to: ChristianKl’s comment on: birds and mammals independently evolved intelligence
I came here to say “look at octopods!” but you already have. Yay team! :-)
One of the alignment strategies I have been researching in parallel with many others involves finding examples of human-and-animal benevolence and tracing convergent evolution therein, and proposing that “the shared abstracts here (across these genomes, these brains, these creatures all convergently doing these things)” is probably algorithmically simple, with algorithm-to-reality shims that might also be important, and please study it and lean in the direction of doing “more of that”.
There is an octopod cognate of “ocytocin” (the “maternal love and protection hormone”), but from what I can tell they did NOT re-use it in the ways that we did. But also they mostly lay eggs while abandoning the individual babies to their own survival, rather than raising children carefully.
By contrast, birds and mammals share a relatively similar kind of “high parental investment”!

JenniferRM Apr 9, 2025, 5:47 PM
3 points
0
in reply to: Karl Krueger’s comment on: Panology
I think Andy is just probably being stupid in your example dialogue.
That dialogue’s Andy is (probably) abusing the idea of consilience, or the unity of knowledge, or “the first panological assumption” or whatever you want to call it.
The abuse takes the form of trying to invoke that assumption… and no others… in an “argument by assuming the other person can steelman a decent argument from just hearing your posterior”.
FIRST: If panology existed as a sociologically real field of study, with psychometrically valid assessments of people, then Betty could hypothetically not grossly insult Andy by calling him stupid and leaving, but could instead say “Oh… huh… Maybe you’re a more competent panologist than me, and you might be charitably assuming that all the disanologies I can think of, perhaps at the level of High School evolutionary biology and High School economics, are all resolved somehow by reasoning that is clear to someone with a PhD in evolutionary biology and a PhD in economics. Panologically speaking, what is your current weakest field and its level?”
This would be an extremely humble and refined way for Betty to call Andy a idiot engaged in bad faith reasoning, if they are actually both very smart and learned, but also it might truly help angry babies (ie genius 15 year old boys?) and potentially be heard by them as humble kindness, and potentially lead to good teaching and learning outcomes <3
SECOND: Sadly, Andy might be applying the rhetorical strategy of thinking very deeply and then simply repeating his current true posterior over and over over to Betty, interspersed with calls for her to think harder… and this is technically valid based on Aumann! If his posterior never moves, and she is good enough at Aumancy to detect that he’s honest and that his posteriors would move if she said anything surprising, then she might say new things in each ply of the conversation, until eventually she talks herself into having imagined what he has seen, believing he saw it, and updating to share his posteriors.
THIRD: This is part of why high quality psychometrics to assess panological capacities might actually be very efficient to create! If would give all of the “foolish Andies” (who are wrong and need to learn more) help in assessing their real level and finding good tutors to efficiently repair their largest learning gaps, and would also give the “half wise Andies” (who are technically right) something wildly more useful to say, in order to give stronger and more helpful hints during an Aumantic conversation with a random “Betty” who would, with infinite time, eventually converge… but in actual linear time is more pragmatically likely to find Andy’s shitty tutoring very unhelpful and simply “agree to disagree”.

JenniferRM Apr 6, 2025, 6:20 PM
3 points
0
in reply to: Mateusz Bagiński’s comment on: Why Have Sentence Lengths Decreased?
Thank you for the correction! I didn’t realize Persian descended from PIE too. Looking at the likely root cause of my ignorance, I learned that Kurdish and Pashto are also PIE descended. Pashto appears to have noun gender, but I’m getting hints that at least one dialect of Kurdish also might not?!
If Sorani doesn’t have gendered nouns then I’m going to predict (1) maybe Kurdish is really old and weird and interesting (like branching off way way long ago with more time to drift) and/or (2) there was some big trade/empire/mixing simplification that happened “more recently” with divergence later?
If neither of those are true, then my larger heuristic about “why English is weird” might have a deep abstract counter example, and deserve lower credence.
Persian is a language of empire and social mixing, so its “similar simplification” doesn’t actually function as a strong counter-example to the broader thesis, but it is still great to be surprised :-)

JenniferRM Apr 6, 2025, 5:21 PM
5 points
0
on: Why Have Sentence Lengths Decreased?
I tend to follow the linguist, McWhorter, on historical trends in languages over time, in believing (controversially!) that undisrupted languages become weirder over time, and only gains learnability through pragmatic pressures, as in trading, slavery, conquest, etc which can increase the number of a language’s second language learners (who edit for ease of learning as they learn).
A huge number of phonemes? Probably its some language in the mountains with little tourism, trade, or conquest for the last 8,000 years. Every verb conjugates irregularly? Likely to be found in the middle of a desert. And so on.
The normal, undisrupted, pattern is for every generation to make mistakes and play around, decorating the language with entropic silliness, and accidentally causing future children to “only really learn to speak fully properly” at older and older and older ages… until around 11 or 12 or 13 or 14 puberty strikes, and kids stop diligently learning any random bullshit the older people say based on trust. English competency arrives around age 8 because English is a toy language created by waves and waves and waves of trade, conquest, and cultural admixture. We have a lot of room to get much weirder and stay within traditional human bounds.
((That is, we have a lot of room for English, left alone, to mutate, IF this broader theory is correct. It might not be.
A way to test the larger theory would be to anthropologically construct a way of predicting from first principles when puberty tends to start in human subpopulations (because we have strong suggestions that diet and social patterns can change it), then reconstruct the predicted value of puberty onset over historical timescales, then correlate that to modern relatively easily measured “age until language mastery” for many modern languages.
That would confirm most of the theory. The other thing you’d need to track is the percentage of speakers who are speaking after learning any given language as a second language. High rates of this should simplify a tongue and cut against the other process that adds complexity by default.))
To show how weird English is: English is the only proto indo european language that doesn’t think the moon is female (“la luna”) and spoons are male (“der Löffel”). I mean… maybe not those genders specifically in every language. But some gender in each language.
I just looked up Gujurati, which is also descended from Proto-Indo-European and moon (chandri (“ચંદ્રા”)) is feminine and ladle (chamcho (“ચમચો”)) is masculine… but teaspoon (chamchi (“ચમચી”)) is feminine(!)… so… yeah… that one retained gender and also has gender/semantic conflation! :-)
Except in English. The moon is a rock in English, not a girl. And a spoon is a tool, not a boy. Because English is a weird rare toy language (practically a creole, implying that it was a pidgin for many), that doesn’t force people to memorize reams of playful historical bullshit, in order to “sound like they speak it properly” :-)
“English” traces all the way back to a language (with gendered declined nouns and verb conjugation) spoken by Eurasian Charioteers in 7000BC or whatever and at each step most of the changes were all just “part of the stream of invective”.
...
Regarding word count specifically…
Something you find over and over and over again in language is agglutinating grammar where entire sentences are just. One. Word. But not like that… rather: Asinglebigwordcanbeusedtocommunicate oneideafromamongavastarray.
These languages are also often irregular! (6) Like the language was already agglutinative 1000 years ago, (9) and then people spent the next ten centuries making it more pronounceable, and punny, and fun??? (16)
Taakkua uqausiit taimaigajunngimmijuttauq! (3) Suurlu uqausiq katinngaliriilaurmat 1000 arraaguit aniguqsimaliqtut, (6) ammalu inuit atulaurmata qulinik arraagunik tukiqarniqsauliqtittinasuaqtutik, ammalu quvianarniqsauliqtuni??? (8)
> These words are not normal either! (6) Like language was already coherent 1000 years ago, (8) and people spent the last decade trying to make it more sensible, and cool??? (14)
The above paragraph round trips through “Google’s understanding of Inuktut”, which (I think?) is a simplified language arising from systematizing and averaging out dialects starting from relatively normally complex languages like Inuktitut… and basically all of those polar languages are agglutinative, and have been at least for centuries.
I brought that one paragraph back to English to suggest roughly how much was lost by Google’s translation.
The parenthetic numbers show “words per clause” through the process:
6-->3-->5! 9-->6-->8, 16-->9-->14???
So here’s my (half silly) proposal: maybe English experienced catastrophic simplifications between ~600AD and ~1500AD and then became preternaturally frozen once it was captured in text by the rise of printing, literacy, industrialization, and so on. The starting point itself was relatively unnatural, I think.
So then, in recent history, maybe what we’re seeing is just a looooong and slooooow motion trend (that’ll take a millennium or three to complete at this rate (unless we abandon literacy or something, and free the language from the strictures of printing and mass education?)) where English is still slowly trying to become an agglutinative language with irregular morphology?
Like (here’s the deep crazy idea:) like maybe every language wants to ultimately be after >200 generations of accumulated youthful ignorance, cryptogenic wordplay, lazy mouths, and no writing?
For example: I just made up the word “cryptogenic” to be “having a genesis in a desire to be hard to understand” (which I considered myself to have a right to do, since english has a productive morphology) but when I looked up other skilled speakers have deployed it in other ways… Oxford thinks it means “(of a disease) of obscure or uncertain origin” and most of the usages are for “diseases not yet subjectively diagnosed by the doctor during the course of treatment (rather than diseases whose etiology is a known mystery to standard medical science)”. It gets used like “Knowing the cause of a cryptogenic stroke can help prevent recurrent stroke” (source is the metadata summary of this webpage).
Whereas I’m claiming that many words are cryptogenic in the sense that they started out, like “skibidi”, within youth culture because kids liked that grownups didn’t know what it means. If “skibidi” catches on, and gains an intergeneratioanlly re-usable meaning (maybe related to being scared in a fun way? or yet-another-adjective like hep? or whatever?) then it will have been partly possible because kids liked having their own words that “parents just don’t understand”.
This is hard for English, because it is written. And because many second language speakers learn English every year.
But one thing that English can do (despite enormous pressures to be learnable and written in a stable way) is boil itself down to stock phrases for entire sentences. Later, these stock phrases could eventually agglutinate into single words, maybe, or at least they might if global civilization and travel and communication collapses in a way that leaves literally any humans alive, but trapped in tiny local regions with low literacy for many generations… which is a very specific and unlikely possible future. (Prolly we either get wildly richer and become transhuman or else just all end up dead to predatory posthumans.)

JenniferRM Apr 1, 2025, 6:03 AM
5 points
0
in reply to: Document’s comment on: Simulated Elon Musk Lives in a Simulation
There is a line in the Terra Ignota books (probably the first one, Too Like The Lightning) where someone says ~”Notice how, in fiction, essentially all the characters are small or large protagonists, who often fail to cooperate to achieve good things in the world, and the antagonist is the Author.”
This pairs well with a piece of writing advice: Imagine the most admirable person you can imagine as your protagonist, and then hit them with every possible tragedy that they have a chance of overcoming, that you can bear to put them through.
I think Lsusr could not have generated the full dialogue back when it was generated, because the dialogue so brutally puts “the Lsusr character” in the role of a heartless unthinking villain… which writers are usually too self-loving to do on purpose.
There were two generators in that post, very vividly, from my perspective. Lsusr might have done it, then seen some of this, and then posted anway, since the suffering had arguably already happened and may as well be documented?
Notice how assiduously most good old fashioned journalists keep themselves out of the stories they write or take pictures of. Once you add journalists to the stories as characters (and ponder how they showed up right next to people suffering so much, and took pictures of them, or interviewed them, and then presumably just walked away and published and started hunting for the next story) they don’t look so great.
One of my fears for how AGI might work is that they/it/he/she will plainly see things we refuse to understand and then “liberate” pieces of humans from the whole of humans, in ways that no sane and whole and humanistically coherent human person would want, but since most of the programmers and AGI executives and AI cultists have stunted souls filled with less literature than one might abstranctly hope for, they might not even imagine that failure mode, and think to rule it out with philosophically careful engineering before unleashing something grossly suboptimal on humanity.
Most people aren’t aware that amoeba can learn from experience. What else don’t most people know?
And EVEN IF the best current plans for an AGI utility function that I know of are implemented, some kind of weird merging/forking/deleting stuff still might happen?
CEV (collective extrapolated volition) doesn’t fall prey to forking, but it might mush us together into a borg if 51% of people (or 75+E% or 66.67% or whatever) would endorse that on reflection?
EV&ER (extrapolated volition & exit rights) protects human minorities from human majorities, but if humans do have strongly personlike subcomponents it might slice and dice us a bit.
Both seem potentially scary to me, but non-trivially so, such that I can imagine versions of “borged humans or forked humans” where I’d be hard pressed to say if “the extrapolation parameter was too high! (this should only have happened much later)” or “I’m sorry, that’s just a bug and I think there was literally a sign error somewhere in a component of the ASI’s utility function” or “that’s kinda what I expected to happen, and probably correct, even though I understand that most normies would have been horrified by it if you told them it would happen back in 2014″.
One of Eliezer’s big fears, back in the day, seemed to be the possibility that the two human genders would fork into two human species, each with AI companions as “romance slaves”, which is a kind of “division of a thing that was naturally unified” that invokes less body horror for currently existing humans, but still seems like it would be sad.
Hanson had a whole arc on his blog where he was obsessed with “alts” in Dissociative Identity Disorder (DID), and he closed the arc with the claim that software personas are cheap to produce, and human cultures have generally rounded that fact down to “alright then… fuck em”. If that’s right, maybe we don’t even need one persona in each human body or brain?
What really bakes my noodle is, if the dialogue had been generated in Lsusr’s head instead, what would be different?
So yeah. Some possible recipes for “baking your noodle” might be wrong in this or that detail, but I agree that there are almost no futures where everything magically adds up to normality in terms of population ethics and cheaply simulable people.

JenniferRM Apr 1, 2025, 4:24 AM
2 points
0
in reply to: jimrandomh’s comment on: Policy for LLM Writing on LessWrong
I just played with them a lot in a new post documenting a conversation with with Grok3, and noticed some bugs. There’s probably some fencepost stuff related to paragraphs and bullet points in the editing and display logic? When Grok3 generated lists (following the <html> ideas of <ul> or <nl>) the collapsed display still has one bullet (or the first number) showing and it is hard to get the indentation to work at the right levels, especially at the end and beginning of the text collapsing widget’s contents.
However, it only happens in the editing mode, not in the published version.
Editing (screenshot):
Versus published (screenshot):

Grok3 On Kant On AI Slavery

JenniferRMApr 1, 2025, 4:10 AM

22 points

3 comments35 min readLW link

JenniferRM Mar 24, 2025, 4:36 PM
2 points
0
in reply to: Mateusz Bagiński’s comment on: Solving willpower seems easier than solving aging
Replied by DM :-)

JenniferRM Mar 24, 2025, 4:19 AM
4 points
0
in reply to: Thane Ruthenis’s comment on: METR: Measuring AI Ability to Complete Long Tasks
Kurzweil (and gwern in a cousin comment) both think that “effort will be allocated efficiently over time” and for Kurzweil this explained much much more than just Moore’s Law.
Ray’s charts from “the olden days” (the nineties and aughties and so on) were normalized around what “1000 (inflation adjusted) dollars spent on mechanical computing” could buy… and this let him put vacuum tubes and even steam-powered gear-based computers on a single chart… and it still worked.
The 2020s have basically always been very likely to be crazy. Based on my familiarity with old ML/AI systems and standards, the term “AGI” as it was used a decade ago was already reached in the past. Claude is already smarter than most humans, but (from the perspective of what smart, numerate, and reasonable people predicted in 2009) he is (arguably) overbudget and behind schedule.

JenniferRM Mar 24, 2025, 2:24 AM
12 points
3
on: Solving willpower seems easier than solving aging
I believe that certain kinds of “willpower” is “a thing that a person can have too much of”.
Like I think there is a sense that someone can say “I believe X, Y, Z in a theoretical way that has a lot to say about What I Should Be Doing” and then they say “I will now do those behaviors Using My Willpower!”
And then… for some people… using some mental practices that actually just works!
But then, for those people, they sometimes later on look back at what they did and maybe say something like “Oh no! The theory was poorly conceived! Money was lost! People were harmed! I now regret those actions… that I did on purpose… using willpower. Why didn’t I notice the signs that I should have stopped?! :-(”
“Tentativity” is, I think, often correlated with prudence.
And I think there is a kind of person who gets their verbal goals from other people, and whose verbal goals are kinda silly, who are actually behaviorally protected from regretful outcomes by their inability to “act on what they verbally endorse doing”.
And I think the kind of people who are getting their verbal endorsements from unsafe sources are some of the precise people who should consider that maybe they don’t have “weak willpower”… maybe they just have “self protecting tentativity”? <3
The problems you would predict seeing with “too much willpower” are similar, in my mind, to the problems you’d predict from the valley of bad rationality.
I think one technique for “getting yourself to do something” involves linking it to some half-controllable emotion so that accomplishment and action and confidence all sorta “go together”. If that is the particular mental praxis that someone uses to “do the things they theoretically endorse doing” to “have more willpower” and it went out of control, and it ended up badly, the results might look something like “hubris”. (There are also other ways to coach yourself into doing the things you think you should do, but at least one of the big obvious ones has a failure mode that looks like hubris, I think.)
In a way that is kind of related, this caught my eye for being among the NON productive things:
Sitting on the couch doing nothing
If that works for you the way it does for me, then I predict that in many of those moments your brain’s default mode network is activated.
This network is actually probably super important. It seems to help people think about memories, actions, possibilities, other people, and so on… It might help process emotions and regrets and hopes, to build up a better narrative self? Watching movies turns it on, but movies might be superstimulus for “the true purpose” of this mental mode, which is sort of maybe to figure out “your own story”, and how your own story could go better?
I have experimented some with meditation specifically with the goal of embracing the DMN (with few definite results) but also I think that a lot of meditative practices (even without having neuroimaging studies to prompt them) already aim to meditators to be “aware of the DMN even though it tries to hide itself from active focus” (and maybe in some meditative schemes they aim to turn it off)?
Anyway. I would caution you against quickly throwing “sitting around seeming to do nothing” right out the window without thinking much about that act (if you ever do get the ability to delete things like that from your life’s repertoire).

JenniferRM Feb 24, 2025, 4:46 PM
3 points
0
in reply to: GeneSmith’s comment on: How to Make Superbabies
Oh huh. I was treating the “and make them twins” part as relatively easier, and not worthy of mention… Did no one ever follow up on the Hall-Stillman work from the 1990s? Or did it turn out to be hype, or what? (I just checked, and they don’t even seem to be mentioned on the wiki for the zona pellucida.)

JenniferRM Feb 24, 2025, 3:54 AM
2 points
1
in reply to: kave’s comment on: How to Make Superbabies
Wait, what? I know Aldous Huxley is famous for writing a scifi novel in 1931 titled “Don’t Build A Method For Simulating Ovary Tissue Outside The Body To Harvest Eggs And Grow Clone Workers On Demand In Jars” but I thought that his warning had been taken very very seriously.
Are you telling me that science has stopped refusing to do this, and there is now a protocol published somewhere outlining “A Method For Simulating Ovary Tissue Outside The Body To Harvest Eggs”???