site banner

Friday Fun Thread for March 31, 2023

Be advised: this thread is not for serious in-depth discussion of weighty topics (we have a link for that), this thread is not for anything Culture War related. This thread is for Fun. You got jokes? Share 'em. You got silly questions? Ask 'em.

Jump in the discussion.

No email address required.

I've flown some long haul in economy, and don't remember ever wanting anything in particular? What kind of thing can the attendants do to improve a flight?

Something you may find interesting: Even within LDS theology, alcohol is not considered evil per se. Right in our Doctrine and Covenants Section 89 (revelations received by Joseph Smith, similar in style somewhat to the Quran as opposed to the Book of Mormon which is a narrative), only "Strong Drink" is specifically called out. And at the time, this was not given by commandment but by suggestion - Joseph Smith, Brigham Young and others famously enjoyed and procured drink on occasion. I believe Brigham even operated a whiskey distillery (tsk tsk). Eventually, modern revelation has declared it verboten - BUT - we also know that Jesus drank wine. He didn't drink grape juice as the well-meaning primary teacher is oft to akwardly suggest, no, he drank wine and he was perfect. Therefore - alcohol is not some out and out evil.

The Word of Wisdom (section 89 as it's called) specifically says it's given in response to "evil designs of conspiring men in the last days" - and at least for me, I agree that even 300 years ago alcohol made sense. You made it yourself, you didn't have psychologists working on mass market ad campaigns to create the most compelling possible product, you didn't have stronger industrial made drinks at absurdly high price points, motor vehicles, social media...The world is different now.

This reminds me of the time I was slamming out code to map a 3d Perlin texture to any arbitrary polygonal model using some alternate coordinate system I'd read about 13 years ago and have long since forgot. I got blackout drunk working on it, but when I woke up on the couch in the morning and compiled what I had, it had no errors and worked flawlessly. I remember feeling really excited, but intimidated, by the problem when I was thinking it through. I recall enough of the night to know the alcohol completely took the intimidation away, and left only the excitement and a devil may care attitude. It's just code, how bad can you fuck it up really?

Consider me a believer in the virtues of alcohol.

I swear I’ve seen a similar argument before on this site. With the focus on Romans as having basically all their social events under a light to moderate buzz.

There are certainly some situations which benefit from ye olde social lubricant. I’d guess socializing and maybe politics are the best cases. Engineering, athletics, and operating heavy machinery…not so much.

That may have been mine, somewhere in my comment’s history. I believe I mentioned that their wine use was usually watered two parts to one alcohol, and their alcohol was perhaps a bit lower ABV than ours. OP’s is much more beautifully written

According to an askhistorians post, just going off of volume consumed significantly overestimates their alcohol consumption -

Most ancient wines probably had an alcohol content around 15% ... In most cases, however, the wine consumed at social gatherings was probably between two-thirds and three-fourths water, which would have reduced the alcoholic content to about that of modern beer.

Your study is probably cherrypicked / poorly conducted / something, just because it's a cool-sounding popular study about human psychology.

Any amount of alcohol temporarily reduces intelligence and precision in your physical movements - a tiny bit if buzzed, a lot if drunk. Having that all the time seems dumb.

Natural selection has tuned our ability to be glad and bond with others over millenia - it'd be weird if the baseline value was simply too low, and constant intake of fermented grain was ideal. Whatever effect alcohol has on various neurotransmitters would be quite 'easy' to evolve ... but it didn't, and you're going to knock all sorts of useful and functional relationships out of balance by being constantly buzzed.

(OP was a good post though!)

Any amount of alcohol temporarily reduces intelligence and precision in your physical movements - a tiny bit if buzzed, a lot if drunk.

Not true, alcohol is considered a PED and is banned in shooting competitions:

I'd heard of that, but that's because (low confidence) shooting accuracy depends on incredibly precise positioning of extremities, and is harmed by nervousness and a strong heartbeat, and from your article

The best known of the drugs used by shooters to relax themselves and potentially slow their heart rate are alcohol

Relaxing and slowing heart rate aren't useful for the overwhelming majority of either sports or day-to-day activities of a roman i'd imagine

Your evolutionary argument is nonsensical. Western civilization evolved alongside alcohol from the very first semblance of high-level social order. So much so, that myths were created around the development of alcohol. Civilizations which used other naturally-derived pharmaceuticals, like saffron, had myths develop around that drug. (We evolved, additionally, alongside the unnatural acts of eating grains and collecting salt from distant places.) If teetotalism were evolutionarily beneficial, then the societies which never spent the enormous resources on alcohol would have triumphed. In fact, they were utterly dominated by the alcoholists. History grants us a sample of thousands of tribes and organizations of human society, and continually the ones who imbibed bubbled to the top of the chain.

Among the reasons why this may be, is that alcohol is a very inexpensive form of pleasure. Instead of having men fight other men or try to fuck all the women or jump off cliffs or whatever, you just give them alcohol and the enjoyment will be proximal. This is especially important when the labor is boring, like “non-evolutionary” farming, versus hunting which is more fun.

There are two ways to look at “evolving”. The one is we should only do what we evolved to do, in which case we just ditch farming. The other is that we evolved to dominate and spread our genes any way we can, in which case we must lay claim to farming and slcohol. Now, maybe it’s the case that the West has new forms of cheap pleasure that replace alcohol. But this doesn’t mean that alcohol wasn’t our reliable help for many thousands of years.

Western civilization evolved alongside alcohol from the very first semblance of high-level social order. So much so, that myths were created around the development of alcohol.

Ancient western civilizations varied a lot in their alcohol consumption. Myths about everything in day-to-day life existed, whether new or old.

If teetotalism were evolutionarily beneficial, then the societies which never spent the enormous resources on alcohol would have triumphed

? Evolution often preserves maladaptive traits, either because selection was too weak, it competed with another effect (for instance: alcohol is a source of calories and a way of preserving them), or some detail of the selection process (individual/genetic and society level selection is a lot slower than culture-spreading selection)

Surely the great Islamic empires were great civilizations, at least during the "Islamic Golden Age"?

Hm, for the wine AskHistorians post, the liter being undiluted does make the wine's later dilution irrelevant. My mistake.

Searching elsewhere for roman wine consumption, from wikipedia:

Phillips estimates that on average, each member of Rome's urban population (man, woman or child) consumed half a litre of undilute wine per capita daily.[1] Tchernia and Van Limbergen estimate the same average consumption levels per diem, per capita throughout the Greco-Roman world.[9]

"Man, woman, and child" at "half a litre per day" is comparable to 1L/day for an adult man, I think? But following that citation, they took estimated wine production for Rome divided by the Roman population - and my sense is that, combined with the % ethanol of ancient wines, isn't necessarily that accurate. But all accounts seem to agree that alcohol consumption was widespread, to the point that during the Empire drunkenness was an issue.

Does it really, though? For almost all social interactions, the ability to be more earnest, more long-suffering, and less constrained by cognitive ruts is an advantage

If you're drinking alcohol consistently, throughout the day, as the post describes, it'll harm your intelligence and physical capacity for the intellectual or physical work you have to do. Said intelligence and subtlety might also be useful for social interactions - which aren't just a way to bond with the bros, but often involve competition, subtlety, and antagonism. Every part of social interaction is 'cognitive', it doesn't really make sense that a general depressant would aid in social interaction.

Throughout human history, almost every civilization that a Westerner would recognize as a civilization has both relied upon grain and developed some form of ubiquitous alcoholic beverage from it.

Eh, according to wikipedia the high levels of alcohol consumption in Rome only took place during the empire - and the rate of consumption significantly differed - a number (that I didn't check) was 1.8L/day/10 men in egypt.

So, are you just going to completely ignore things like bar fights and other obvious examples of how alcohol consumption often substantially increases men’s propensity to interpersonal violence? “Promoting collective action, good will, and forgiveness between strangers” might be the way that some populations and individuals respond to alcohol consumption, but many others find that it exacerbates violence and dysfunction.

The problem, though, is that what for one man is a moderate amount of drinking might be, for another man, enough to get him drunk enough to be a problem for himself and for others. And it’s not always clear to any particular person what the effects of a particular round of drinking will be; I’ve had outings where I found myself significantly drunker than I expected, because of situationally-contingent factors - what I’d eaten that day, the beginnings of a cold/flu that hadn’t begun to manifest when I woke up that day, etc. - and additionally a great many people are terrible judges of their own current state of intoxication.

We’ve all met plenty of people who, while visibly and obviously intoxicated, insist that they are sober enough to drive, or sober enough to take another shot. Being intoxicated, since it is a euphoric sensation, naturally incentivizes the consumption of more of the intoxicant in order to prolong and amplify that sensation. Sure, very experienced drinkers with a strong sense of their own tolerance - in addition to a strong penchant for self-control - can recognize signs of drunkenness in real-time and abstain from further consumption; I think that you’re significantly over-estimating the percentage of the population that fits that description, and under-estimating the percentage that get drunk without intending to and cause all sorts of problems.

More comments

How Rationalist (community) Are You survey. Kind of curious how y'all rank. Its an Aella survey of course.

I got "Highrat: 95-99th percentile" I think this is because I like all the blogs, but apparently there are some counterintuitive questions that I took the "rat" side of

Your score is 18.59!

Highrat: 95-99th percentile: A rare form of rat, you almost definitely are a programmer with lots of controversial views and a brain that's almost entirely boy. You know the AI alignment arguments both ways. You may not have been to a CFAR workshop due to it being too social.

"You are Normie: 30-50th percentile"

Score: -7.56

I've never felt so proud to be lumped in with the normies. I'm guessing it's because I'm not a science nerd, I don't read navel gazing inside baseball EA blogs, and I'm a socialcon. I might have scored differently had different questions been asked.

5 - 15th percentile.

Unratlike, 5-15th percentile: Low! You don't jive with most of rationalists say, AI definitely isn't going to end the world, you probably have a great exercise routine.

I don’t like blogs, I really can’t stand pages long drawn out waffling and I’ve long since come to the conclusion that being talked about here often is a major anti-recommendation for an author as far as I’m concerned. I also suspect the nanotech, condensed matter physics and genetic engineering questions assume ”rationalists” to have either studied the domain academically or (more likely) have major Dunning - Kruger about them.

Edit: I suspect the questionnaire could have been reduced to just a single question while keeping much the same result (albeit at much cruder scale): ”Do you like intellectual masturbation even if it doesn’t benefit your professional life or major hobbies?”

I got -5.38, which apparently makes me

Toe-in Rat, 50-70th percentile. You probably play board games, know what prediction markets are, and maybe occasionally read a blog or two. Not really a full-fledged rationalist, but their world is next door.

I suppose I do occasionally read a blog or two. I don't have any interest in board games. I know what prediction markets are, but don't have any particular desire to participate.

I got 70-85th percentile, which is… odd, because I definitely don’t consider myself to be a Rationalist (but I am a terminally-online nerd, so, if that’s the level of granularity you want to work at…)

I answered disagree/strongly disagree to the “are you really smart and have you read a lot of books?” questions - I wonder if that decreased the score or increased the score.

75p light rat, I was wishing for it to be lower. I strongly disagreed to that poem and the physics and the nanotechnology question as well.

What’s wrong with the rats?

I don't like their Aesthetics. And on a personality level, they tend to be too neurotic.

I feel like could get along with Robin Hanson and Bryan Caplan not with Yudkowsky, Scott Alexander, and Scott Aaronson; the latter 3 are too neurotic.

Lightrat. I wonder if being annoyed at every question being "agree/disagree" correlates with the result.

30-50th percentile, normie. Yeah that's about right. I'm not exactly a normie but I've never considered myself a rationalist. I'm just here because I used to like Scott's insights on culture war topics, and because I still enjoy the insights here.

Fun thing, this question threw me for a loop:

Socially speaking, you tend to be more

Social leftists tend to prefer lower government involvement in social issues, for example allowing drugs and abortions. Social rightists tend to prefer higher government involvement in social issues, for example outlawing sex work or obscenities.

I would prefer a lower socially mandated conformity (not necessarily via the government) on social issues, which happens to favor the "left" side currently. Like, on the 2d political compass I'd be left-libertarian, but absolutely not left-totalitarian.

Highrat: 95-99th percentile: A rare form of rat, you almost definitely are a programmer with lots of controversial views and a brain that's almost entirely boy. You know the AI alignment arguments both ways. You may not have been to a CFAR workshop due to it being too social.

cool, i guess

I got -14.53

Normie, 15-30th percentile: You're not that antirat, but you're definitely not a rationalist either. Maybe you have a degree in literature, or have a few dogs, or like watching sports.

I got TrueRat, 85-95th percentile. I am also creeped out by how closely I was pegged by the flavor text…

No question, you are One Of Them. Probably you work at a rationalist-org, or have seriously considered it. You probably have a whiteboard in your bedroom.

Not surprised. I’m pretty deep in the culture but have some fundamental disagreements with a lot of the nerdier stuff.

High rat (95-99).

The flavor text saying I'm "almost definitely...a programmer" is a little funny since I am definitely no such thing. I do think the apparent overlap between computer science types and philosophy/politics/economics circles is interesting, though.

Toe-in Rat. Sounds about right.

I got 95-99th myself, to my utter lack of surprise!

Lightrat. I'm surprised I scored so high, to be quite honest. But I'm curious how people said "studied" for different things. Studied advanced physics like, at a university course you passed, or like on a youtube playlist from a random guy with crazy eyes? But then, maybe the Rats who didn't actually study physics thinking they study physics is what makes them rationalists.

oh, I was a bit confused by those questions, I answered yes for the physics ones because I took grad courses in those subjects. Was the expectation that rats would say they "know how to design nanotechnology" from just reading the sequences or something?

I'm highly skeptical anybody answering that survey actually knows how to design nanotechnology.

Maybe the question should've been "have you read Drexler's Nanosystems and understood all the math?", because "nanotechnology" isn't exactly as well defined as something like quantum mechanics.

Like someone mentioned elsewhere in the thread, I think those could have been there with intention of capturing a Dunning Kruger effect. Which I think is actually and underused tool in surveys to capture certain overconfident opinions people might have.

If graduate school doesn't count as studying something, I don't know what does.

The question isn't necessarily expecting overconfident claims, though. It just expects really high chance of comorbidity.

It's tough maybe being a girl or something.

We're just too pretty for the other rationalists, Fruck.

I got MidRat, which surprised me. I didn't think I'd rank that high, being an evangelical Christian who is only mildly worried about AI. I'm not sure what pushed me over the edge: being able to recite a poem from memory? Having a good opinion of Mormons? I'd love to see the correlations.

The test seems to assume that a lot of things are exclusive to Rats that aren't.

Like with the little blurb at the end with the toe-in rat score seems to imply that prediction markets are a Rat thing rather than a polite version of an assassination market.

I couldn't do the poem one, so I don't think it's that.

yep, got a 25.38 score, which makes me HighRat too

Lightrat 75%. I'm surprised it's that high.

She really pegged me:

Toe-in Rat, 50-70th percentile. You probably play board games, know what prediction markets are, and maybe occasionally read a blog or two. Not really a full-fledged rationalist, but their world is next door.

I host a Saturday board games event monthly at church, know of but have never visited a prediction market site, and peruse blog posts but don’t follow blogs.

5-15th percentile. I’m a little surprised I’m THAT low, but yeah, I’m only a semi rationalist.

The question that equates economic leftist and economic liberalism and calls economic liberalism economic rightism hurts me.

Why did Americans have to ruin the word liberal.

Yeah amen. I went liberal on the economic question before I realised what she meant with the social, although I don't know how it affected my score.

I asked Bing AI to help me make a Blood on the Clocktower character, here's the result:

It's an actually interesting character, I discussed it with the pals and they thought that it was quite overpowered if anything.

Also it was a flash in the pan, it took me a while to convince the AI to help me (it kept insisting that it was not a game designer for some reason), then I got this, then I got about a dozen of nonsense/boring suggestions.

On a related note, come play with us in our Blood on the Clocktower discord!

It's a variation on Mafia/Werewolf but with several important distinctions that make it superior, and especially superior for internet games, and even more superior for text games with 24h/game day (but we also play voice games sometimes btw!).

First of all, everyone gets a character with an ability. Abilities are designed to be interesting and include stuff like "if you die in the night, choose a player, you learn their character". Second, dead players' characters are not announced, they can still talk with the living, and retain one last ghost vote, so if you get killed you're still fully in the game and maybe even more trusted. So you get games where everyone is engaged from the very start--because you want to privately claim your character, maybe as one of three possibilities, to some people--to the very end when you cast your ghost vote for who you think is the demon.

Lately we had some rdrama people join (including Carp himself!) so it would be nice to balance their deviousness and social reads with having more themotte folks. We were historically very balanced:

My favorite voice game (not our group, but we have had similar shit going down):

I want to reiterate that it's worth trying a game if you're at all interested. You can dip your toe in and play a traveler who can come and go at any time or just read some of the archived games to get an idea. We've onboarded dozens of players with a wide variance of interests and people are constantly needing to explain niche game mechanics to each other so it's really no big deal if you don't know anything before joining.

the games are wacky and fun! whether you're someone like @carpathianflorist who will trust someone for no reason at all, whether you're like @aqouta who wants to build charts to figure out who the demon is, or like me who will get into a counterclaim with my own minion my first game, there's a lot of fun to be had!

I am the smartest

It’s a very fun game, can confirm. Also that character is interesting but OP as hell! Maybe like, rewind one action once.

It's not really OP in my opinion. It's sort of like a gimped Professor: resurrects a player but only the last executee. And like the Professor if it's out the Demon can just kill him. But on the other hand you get the whole day of info about who nominated who and who voted for who, so that could be incredibly strong.

It's strength will really depend on some key rulings in edge cases. The interesting bits are if you blow it to do something like get an extra savant read but the savant ends up being a minion just pretending

i think the thing too with is that for savant an ST can just give them the same savant bits if seems too overpowered. it's definitely useful though and i think it could be very useful.

The ST can give you the same Savant info, welcome to the Groundhog Day. Fortune Teller and the like which get to choose what info they get, get a bit OP. On the other had, the evil get to redo their actions too, in the light of what's revealed. Or kill the Timekeeper if it's too scary.

The main problem with it as I see it is how dependent it is on synergies. Most other characters in the game work alone somewhat even if they also have strong synergies. This ability is basically useless unless it's used to accelerate something else.

Hi guys!

Someone suggested we recruit for the motte/drama combined BotC game here. I figured I'd use this moment to spoil this idea before someone far more trustworthy and reputationally sound actually recruits more people with propensity towards studying autism charts. Don't do it!..

The next game will be our 30th game! It's a fun little community for arguing about things that are definitely inconsequential for a change.

For those of us lucky enough to have access to ChatGPT, what are your favorite plugins?

Which ones do you think are the coolest? I’m lining QChatGPT, which helps with QGIS maps.

S A T O R 

A R E P O 

T E N E T 

O P E R A 


This is an interesting palindrome from the first century. It is called the Sator Square. It found its way eventually to every European country, as well North Africa. Every word in this Palindrome is a common Latin word when read either left to right or right to left, with the exception of the backwards reading of opera which reads as arepo, having no other mention in Latin but may be construed as a proper name. Uniquely, this Palindrome is not just left to right and right to left, but up and down and down to up.

Its text means, when read as a book, “the farmer, Arepo, masters the wheel with effort”, or “Arepo the farmer holds to the work wheel”, or something along these lines. Speaking of lines, it may be read in the boustrophedon way, in alternative direction, like how a farmer plows his field, which is fitting and common in old Roman inscriptions, and this may render the meaning as “the sower sustains the works, the sower keeps the works”, though this fails as a meaning-based palindrome when read backwards. This is akin to “as you reap, so you sow”, which has a long-standing mythical and spiritual meaning for the Eleusinian Mysteries and other allegorical Roman stories and rituals, and later Christianity which synthesized many of these allegories.

Historians have debated what it means and its purpose, but historians are also dumb, and frequently ignore the forest by analyzing the trees. The fact that this is a four-way Palindrome, and that it has a normal and mundane meaning in Latin, and that it is applicable to the majority who worked in agriculture while having a vaguely moral message, is all the meaning necessary to explain its wide use and transference. Asking for Arepo to also hold meaning is asking too much, as he’s already holding the wheel.

An interesting discovery was made in the 20th century. A rearrangement of the square leads to an intersecting “cross” of the phrase Pater Noster, united by the N in the middle, with two alphas and two omegas as a remainder. This is certainly just an accidental fun additional Christological reading. There’s no evidence that our ancestors knew or conceived of this in previous centuries.

Because historians are dumb, they cannot accept that the four-way palindrome was popularized and utilized by virtue of being a really good four-way palindrome. They are looking for additional reason for no reason. Humans love patterns. Reading this palindrome square gives my mind a little intellectual orgasm every time I look at it. The surprise at the quad-directionality of the square every time I see it gives me a bit of pleasure. So it did to the ancients. No additional reason is required, and in fact an additional meaning is statistically improbable, bordering on impossible, for a cute symbol that is already packed with meaning and pattern. I’m reminded of listening and hearing Bach, and music theorists see all the wacky patterns he created in his works as kinds of musical puzzles, and they ask “what greater numerological meaning does it have?” It does not, as the puzzle is the meaning.

Speaking of lines, it may be read in the boustrophedon way, in alternative direction, like how a farmer plows his field

Overall fascinating, 10/10 funpost -- but that's not how you plow a field. You want to scoop the dirt into the furrow from your last pass, so reversing directions everytime is the opposite of what to do.

If you start at the edges, you do either end first, working your way towards the middle -- where there is a 'dead furrow', a low spot from scooping a single furrow of dirt outwards on two opposite passes. (which lines up perfectly if you are an awesome ploughman)

Or you can start in the middle piling two furrows on top of undisturbed soil, and work your way to the edges leaving a dead furrow there.

Done this way it would read either "Rotas Rotas Opera Opera Tenet" or "Tenet Arepo Arepo Sator Sator". Nice rhythm; not sure how meaningful.

(How do you say "Pedantic farmer, signing off" in Latin?)

I think the term boustrophedon is older than a mouldboard plough. The Greeks used ards, which allow you to plough in any direction.

That is fair enough -- I think the Romans were first to come up with something like a moldboard?

So I suggest to sub "Ancient Greek farmer" for "farmer".

I wrote a quick and shitty script to bruteforce English Sator squares from a big wordlist. The only one that makes any sort of limited sense is this:

t i m i d

i r a n i

m a d a m

i n a r i

d i m i t

I.e. "[the] timid Iranian madam that Inari dimit(=had dismissed)".

Edit: Of course somebody else had done the exact same thing and found the exact same square before. Oh well.

I tried to bruteforce a Russian Sator square, and got 634 variants (317 when you exclude ROTAS). At least one makes sense:






We're destroying with our minds the hovel for the imam of tinsels. I could also try and add some 3+2 and 1+4 word combos to make more, but most squares already don't make any syntactic, let alone semantic sense.

You could try a big namelist and allow one of the lines to be a name, like the original square

Surprised GPT-4 couldn’t do this. Rip foom.

As it goes with all children. Alas.

Also, had no idea you could use emojis here. Interesting… 🤔

Oh, this takes me back. My great grandma was quite religious, but not very discerning in her tastes, so she subscribed to "Science and Religion", a magazine that was full of esoteric stuff like Sator square. I remember that I loved leafing through them every summer, even though that was more than a quarter century ago.

I wonder why no one has come up with a better four-way palindrome since then. Can someone ask GPT-4 to make one?

Also, the website of the National Puzzlers' League is delightfully Web 1.0.

GPT-4 gave me this after a few tries:

“Creating a 5x5 word square with proper meanings in English is quite challenging, and there aren't many known examples. However, I can provide you with a 5x5 Latin word square, similar to the Sator Square:






In this square, the words are:

REGIS (of a king)

EVENI (to come forth)

GIVER (the giver)

INERT (inactive)

SITER (they may be)

These words have proper meanings in Latin. Keep in mind that, like the Sator Square, the meanings may not form a coherent sentence, but the individual words have meanings.”

Pretty awesome that it parsed my meaning, said hey that’s hard, but here’s something in another language. This stuff is damn impressive.

The actual result is not actually that impressive, though. "giver" is plain Germanic English, "inert" is of Latin origin but anglicised (in Latin it would be iners, with the -t- in the stem only appearing in some flexed forms), and "siter" is not any sort of valid Latin form to my best knowledge ("they may be" is sint), though it's interesting that it leveraged some correct knowledge that si(t)- occurs in subjunctive forms of "to be" in Latin.

Ahh good to know. Yeah still a long way to go with these models. I’m surprised at all the hype given their frequent hallucinations.

For what it's worth, I think the hype is completely justified, and the concern over hallucinations is driven by a combination of the usual motivated reasoning trying to defend human uniqueness, and not understanding what problem has actually been solved. The LLM's unfiltered train of thought should not be compared to a human's well-considered sayings/writings, but to a human's unfiltered train of thought, and I'd be extremely surprised if anyone out there has an inner monologue that is as coherent and intelligent as GPT-4's output. The problem of letting an LLM "think before opening its mouth" has not quite been (publicly) solved yet; but, crucially, there is no obvious reason to believe that the remaining gap (from being able to have a coherent, insightful inner monologue, to shaping the monologue to consider when and what to speak and thereby achieve correctness) is at all difficult to close. We have moderate success with teaching this to human bullshit artists who say the first thing that comes to mind and then make up rationalisations for it after the fact, too.

I like the way you put that, and it’s not something I’ve considered before. What sort of timeline do you have for AGI?

Depends on the definition you use. "Better than the (pointwise) median human at almost every fair test of intellectual ability" (that is, you can't ask it subtle rarely-discussed details of growing up as a meatbag and declare victory when it can't), I'd put at approx. 5 years with a 90% confidence interval of 0 (that is, someone already has built it) to 15, conditional on no significant change to the conditions of development (say, full-scale CN-US war, or a somewhat successfully prosecuted moratorium on research). "(pointwise) better than every living human at every such test" at approx. 20, 90% confidence interval 5 to 50, again conditioned on the same. Caveat for the latter is that I think that this conditioning actually cuts out a lot of the probability space; I'm thinking maybe 75% that something I would count as a significant change happening before we get AGI (second definition).

Interesting. Most people seem to think once we get AGI we’ll speed to super intelligence very quickly, why the 15 year gap? You don’t believe in self recursive improvement?

More comments

Yes, I'm really glad to see someone else point this out! One thing that's interesting about LLMs is that there's literally no way for them to pause and consider anything - they do the same calculations and output words at exactly the same rate no matter how easy or hard a question you ask them. If a human is shown a math puzzle on a flashcard and is forced to respond immediately, the human generally wouldn't do well either. I do like the idea of training these models to have some "private" thoughts (which the devs would still be able to see, but which wouldn't count as output) so they can mull over a tough problem, just like how my inner monologue works.

You can kinda do this in chatGPT - ask a question as a chain-of-thought prompt, then a follow up asking it to extract the answer from the above.

Experimenting with giving ChatGPT-4 a more structured memory is easy enough to do that individuals are trying it out: I find his estimate of AGI-in-18-months a little optimistic, but I can't completely rule out the possibility that the "hard part" of AGI is already present in these LLMs and the remainder is just giving them a few more cognitive tools. We're already so far down the rabbit hole.

This reminds me of "blank palindromes", a gag one Russian writer came up with. Blank verse looks like verse, but doesn't rhyme, so blank palindromes look like palindromes, but don't read the same when reversed. This is a blank word square.

Spoonerisms are really common in Finnish, so people have also played around long time with what one might term "blank Spoonerisms" according to this schema, as well.

Yep I called it out and it said:

Larger palindromic word squares, such as 4x4 or 5x5, with proper meanings in English or Latin are very rare and difficult to construct due to the constraints of the palindrome property and the limited availability of suitable words. The Sator Square remains a unique and famous example of a 5x5 palindromic word square.

Ah, I see you've watched the movie.

Great movie! Gets a lot of flak for being high concept but I love it.