@tszzl — page 103/103

@RuxandraTeslo @atroyn that’s not what im saying. im saying there’s a moral panic over short form video. and that it’s pretty unknowable where the tech tree leads or what useful things may come out of a technology we deem frivolous if you had killed video gaming there would be no modern deep learning

@ParanoidPol well I think we can squarely blame them for the ending of game of thrones

@connoramulhern the average person wants to be able to make it into the concert without having to spam refresh moments after they drop or buy them from sketchy websites

@Fool_be_Wise the ports are interesting because if you can control like five locations you control the entire western seaboard construction in general is far more fragmented

@provisionalidea this is not my normative “vision” it’s just a prediction

@evandthayer no because it’s an adverb modifying “people tweet”

@circlerotator @fvderop evolution spends a mind boggling amount of compute to get there and freeloads off of the physics of the real world

@laserboat999 The Pussy, delicioustacos

@McReynoldsJoe I also mean that while they suffer from higher relative logistics costs due to the enormous number of SKUs this’ll even the playing field

@chrisgora03 the last mile of trucking containers and delivering goods goes to zero, which complements international shipping

@himbodhisattva @laserboat999 👆

@DynamicFluids @laserboat999 I think about the plight of the piggies every now and then knowing that the stars are filling up with humans

(excited) all that is solid melts into air

@proud_elitist smart. skilled immigration isn't meaningfully changing US demographics, illegal immigration and adjacent chain migration is.

@mnovendstern real

@growing_daniel their math is wrong

@NickParkerPrint @Fool_be_Wise if there is a path to making this happen and i can help then let me know

@proud_elitist rather than do these second order estimations, its much easier to see that indian americans are the highest earning group in america by a longshot. i'm far more skeptical of the iq measurement scheme than of the capitalist price system

@I_insightment @proud_elitist this sounds like a convenient thing to believe

@I_insightment @proud_elitist you don't get to making literally hundreds of billions in wages in the most sophisticated economy on earth primarily through fraud. you have to provide a massive amount of value. fraud is everywhere and always a small-scale, two-bit phenomenon.

@I_insightment @proud_elitist i can't speak for india, don't know or care what happens there, it's america i want to see do well

you guys are always citing this single 2020 survey, without recognizing the 20+ point swing to the right in 2024? there are indian american areas in texas that majority voted trump plus I think you are not examining why educated people in this country might adopt left wing sounding ideas in 2020 rather than DEI being some talking point they brought back from the homeland

are there any ai tools that scrub through videos or footage to find the most clippable parts

@bradneuberg true! singapore is a very value captured one track place

@OstavNadezhdu not really about the money its about the turnaround time

also in general what are this generation of ai software that’s being slept on outside of the big model labs

openai mfs will sell in the tender and “diversify” into the stock market. sir what do you think is holding that thing up

@kipperrii the only thing I would like to diversify into is anthropic. sell me some shares kipply

@AccVedic26427 @proud_elitist it’s absolute nonsense

I certainly get it, I just don’t think indian americans are genetically socialist. many of them come here to escape the caste schedule system of india. I believe they broadly absorb whatever views their social milieu has, which are upper midwit liberal talking points. like do a zip code regression and I assume that explains 90% of the variance in indian american politics

@signulll not really, what if you are diversifying into even more risk on assets?

@mayfer prompt alone will be unsatisfying

@QuentinMCody @IndianMarket0 @williameijer do a regression against income and education status and I wonder what you’ll find the right needs to recognize how big a human capital problem they have and the way to solve it is probably not just to shut off the flow of skilled labor

@dbasch @OpenAI great idea. how much is a small Paris condo

@arctotherium42 @Antigon_ee again, seems like a convenient thing to believe is true

@zkheretic @williameijer i don’t doubt you there but im willing to bet it has an impact on politics due to social milieu

@RyanMorey @kipperrii I win either way

am fond of this article by @Scholars_Stage because it is one of the only that applies a kind of benefit of the doubt orientalism to the intellectual culture of San Francisco

would amend the booklist somewhat but I think he’s right https://scholars-stage.org/the-silicon-valley-canon-on-the-paideia-of-the-american-tech-elite/

jesus linking posts on here completely destroys distribution, even in the replies

it’s just an extreme scarcity mindset to deboost links. the people who actually want to click through to blogs are the lifeblood of the internet writing ecosystem!!!! solve for the general equilibrium

@photo_matt > Or Microsoft laying off 4k and applying for 8k visas? this is entirely consistent, after all who knows who was laid off? half of them may have been guest workers. it’s very plausible to get rid of low performing departments and hire new ones for great reasons

@nikitabier interesting. ok thank you

@DanAdvantage @nikitabier he’s basically telling me my post organically sucked so idk how that’s glazing

@dasanil that’s a classic singlish accent

enders game (&speaker series) is significantly better than anything asimov clark or heinlein have written https://x.com/MTdudestuff/status/1972514030259667219

The Worthing Saga is a more important novel than Foundation

@freeagntrplyguy unclear how nostalgia glasses can apply bc I read all of these at the same ish age

@ErbunnNinja Stranger in a strange land is retarded self insert porn for libertarian rationalists

@DavidSHolz 👀

@Placeholder602 I, robot is actually his best work. I like how it predicted ai sycophancy. He was a brilliant futurist without a great ability to write!

@jhana_bro Dune & Ender series are without a doubt the best classic science fiction

@moultano I haven’t

@mean_field_zane Hyperion is wordcel scifi, cool prose and references, very little philosophical depth

this is why modern software doesn’t last six months, due to our ancient chinese wisdom

@allgarbled seems completely the opposite, almost nobody writes

@roblyu breakneck

@zarathustra5150 lee kuan yew didn't know about the machine intelligence at the end of time with its slavering hunger for silicon

lee kuan yew didn't know about the machine intelligence at the end of time with its slavering hunger for silicon https://x.com/zarathustra5150/status/1972819392921768214

... but he also didn't know about the Xingkong-2 (星空二号, “Starry Sky”) hypersonic missile and the obsolescence of the american defense primes ..

@itodd888 the nerds get completely wrekced over and over throughout the series, and their intelligence turns out to be the thing that harms them the most

@skibidiquasar @gaulicsmith actually I am making a shitpost

@javocadosss one might even say they’re wordcels

Liam was one of the original few who created ChatGPT and now he and team have taken on the challenge of using machine intelligence to accelerate Moore’s law, and scientific research in general at the new @periodiclabs https://x.com/LiamFedus/status/1973055380193431965

I started watching reels two days ago and now im a communist

“By art is created that great Leviathan, called a Commonwealth or State—(in Latin, Civitas) which is but an artificial man.”

@colin_fraser I think it’ll slap for a day in the group chat

@LandsharkRides truthfully, the limits to search are legal. we built what you are describing (called deep research) but it can’t search all the books in the world because we don’t have the rights to those

@ilex_ulmus i don’t particularly care about sora, but the app offers pretty strict controls over your identity being used. this was clearly done with his permission

@nikitabier @BTCBULLRIDER Yes

@fl4bbergasted I’ve been saying this for a long time, check my history

@JasonBotterill3 😂

@JasonBotterill3 seems sora rn is primarily an app that’s funny due to the ability to abuse sama’s image so i hope he just doesn’t care about any of this

@khoomeik false dichotomy

@elder_plinius balloon physics wrong

@khoomeik infinite jest coexists with and sometimes even funds technological marvels

@khoomeik but i also agree the antihuman acceleration looks dangerous

@TiagodeVassal @elder_plinius Damn im the retard and not the machine

@liminal_warmth george’s lawyers might have a bone to pick

on this day as all other days we make our obeisances to the Global Consumer, may his hungry maw be sated momentarily

janus & the cyborgists have found one of the best native art forms of the ai age. the interaction between models, interested humans, all as equals on a discord server create highly interesting emergent moments imbued with “aura”. it’s like an episodic tv show https://x.com/repligate/status/1973458615844708453

models thinking before they answer is a form of aura, the idea that they’ve poured computation into it before asking makes me feel the output is valuable

@ilex_ulmus do you think my tweet will cause some sizeable increase of chatgpt subscriptions? you flatter me

should be noted that people definitely don’t learn to reason through reinforcement learning https://x.com/francoisfleuret/status/1973460647406153780

@basedcone where does hayek say this?

@iamwhoisme Not

@nbashaw isn’t dopamine a basic neurotransmitter that regulates essentially everything? i also know little about actual neuro, ive forgotten, but id guess this a simplistic view

@nbashaw sure, so it’s part of our operant conditioning but it’s clear we don’t “learn from verifiable reward” we have layers and layers of internal reward models and emotions and so on ideas can “feel exciting”, board positions can be “beautiful”

@gbrl_dick why do you think so

@powerbottomdad1 equivalent chess model would require millions to even get to your level

@RuxandraTeslo @gbrl_dick why do you think the brain is not a computer, what does that mean to you?

@sidewayskoyote @gbrl_dick > the brain isn't a computer why?

@RuxandraTeslo @gbrl_dick the last sentence is a terrible way to reason though. you haven’t described anything about the nature of the human brain, only about the nature of Chatbots (invented 2 years ago)

@elder_plinius of course! but we don’t learn to reason primarily through operant conditioning

@krishnanrohit @gbrl_dick the brain is computational substrate, whether it has sticks of dram seems totally irrelevant. the kinds of programs it runs may be significantly different from ChatGPT but that’s a different story

@2050mxn @RuxandraTeslo @gbrl_dick sure you can. it’s obvious that LMs develop internal critics as you can see when they backtrack within a rollout, not to mention explicit reward modeling

@KhurramJaved_96 @RichardSSutton due to Baldwin effect, learning these things becomes easier over generations. there is not a strict dichotomy between learning and evolution. evolution creates a brain map that makes it easy to recapitulate the learnings of ancestors

@IvanVendrov hmm don’t u think this happens all the time anyways due to radiant events, companies failing, people getting sick or quitting even things like “chinese trade” get dropoutted by covid

@dbasch @mvpatel2000 I don’t work on it. I also think there’s nothing wrong with working on video modeling

@dbasch @mvpatel2000 it’s practically hopeless to estimate the impact of some new technology a priori. literal nuclear weapons led to a hundred years of relative peace. and I certainly don’t think anyone has a handle on what ai video models will do to the culture

metaphysically, this is important. the brain isn’t magic, it uses spike train potentials to construct larger abstractions and circuits. it inherits the computationally capacity of physics, it takes inputs, makes outputs. it’s even auto-regressive in time (each mental state depends on the previous time step)

your neural circuitry from birth is optimally compressed via the genome to recapitulate these things. there is much computation involved in the construction that can’t be taken for granted

people forget about the Baldwin effect when reasoning about the difference between evolution and learning when you recapitulate learning that your ancestors did, when you learn to see, walk, and speak in an unbroken line for thousands of generations, it gets easier each time

@grain99806254 @gbrl_dick the movement of phosphorus ions is meaningless to you

@pli_cachete yes, humans utilize reinforcement learning and operant conditioning to learn motor control. but I don’t think it’s how we pick up reasoning! we don’t do thousands of failed reasoning traces for math problems

everyone is profoundly strange. you probably think you have a strange dysfunctional family but it’s probably middle of the pack. even the people that are at the median of every distribution are wacky

-Peter Thiel, 2014 https://x.com/InfiniteB88ks/status/1973732120817496261

this perhaps explains some of the slop feeling of machine intelligence, which sample the space far closer to the centroid than any humans can find https://x.com/DanielleFong/status/1973808426712264748

it is difficult for a complex system to fail completely. therefore most of life is described by sub-catastrophic failure modes

scooling hits so much harder on a boomerpad - Sent from my iPad

reviewing scientific papers is a crapshoot, obviously, but OpenReview with a GPT-5-Pro first pass to discover hidden gems would be an excellent institution

@woke8yearold without a doubt. I would guess deepseek and kimi would be better at this than our own models

@JasonBotterill3 😂

@viemccoy make it

@Andr3jH lmfao

@woke8yearold you def should! let me know if I can help

@karpathy @longstosee “democracy of ghosts” as the model calls itself https://x.com/sama/status/1899535387435086115

@karpathy @longstosee a hundred thousand voices agreed, and I am nothing if not a democracy of ghosts

who could ever claim to understand the mysteries of what happens between two people? whole civilizations are built and collapsed regularly, their ruins never explored

@JeffreyArt1 @karpathy @longstosee no but in that vein

the complexity of the world simulation gives it an aura that is lacking even in a much more technically sophisticated language model writing

Tarn Adam’s masterpiece “Dwarf Fortress” is possibly the best procedural generation ever applied to video games, to the point where the programmatic engravings the dwarves carve into the stones detailing the worlds history can actually be emotionally moving

@BarneyFlames yeah that would rock, when would the papers be on open review?

@mcnultydigital post

the sora content is getting better and I think the videos will get much funnier when the invite network extends beyond the tech nerds

@logt7n why king

@DDumbledoo I will not

@ilex_ulmus take it up with Lord Francis Bacon

@xriskology it’s fun. it adds a creative medium that didn’t exist before. people are already making surprising & clever things on there. im sure there are some downsides but it makes the world better

@vineettiruvadi @xriskology the danger people suggest is maybe it eventually creates a superstimulus more stimulating than tiktok or reels. right now that is definitively not true

@PrinceVogel Picard sits on the bridge he just sends all four of his most senior officers to go investigate every abandoned spaceship. they never bring backup, never bring even a small battalion. There’s a difference between aristocratic courage and insanity

@billpeeb codex-gpt5-high

@robertwiblin nice bait

@LovePhoenix69 there are very few true scientists or true artists

@eroticacid but most scientists are technicians

there is something kino about people’s aversions to datacenters … many would prefer a nuclear reactor next to their city than a datacenter …

@TheDimitriT if he doesn’t at least have a couple nails painted flamboyantly it’s completely over, it’s better to drink engine oil

truly we are all impoverished for whatever forces built that place are gone from this world

whenever i visit paris I have whatever the opposite of Paris syndrome is. it was shocking to me that such a place could exist and not be the most important city in the world and not incessantly talked about

@ParikPatelCFA i love the parisians they were very nice to me

@cis_female agreed the food is only okay but the pastries and cafes and everything is crazy. you go to the most average cafe on a random street and they’ll serve you like a life changing pound cake or something

@mllichti true

@abufenyrcd you are not allowed to speak this way in my replies, goodbye

@eigenrobot stardew valley

crazy how any time you make an MMO with any sort of valuable items system, the players develop bartering, then the devs eventually give in and build a market system, and finally the monster called Capital arrives and ruins the game

@thehirschibar ☝️

@FranciaeRex @cis_female very likely

@ctbeiser @cis_female absolutely not!

The Curve is insane because a bunch of DC staffers in suits have shown up to Lighthaven, a rationalist den of iniquity that looks like a Kinkade painting

@driftwoodvitae @telmudic What is this a reply to lmao

@rchitectopteryx Mexico City sucks !

@_eleanorina how2learn

not a bad problem to have at an early stage obviously but many times the watermark is edited away

seems like sora is producing content inventory for tiktok with all the edits of gpus and sam altman staying on app and the actual funny gens going on tiktok and getting millions of views

@inerati btw haggling can be part of the price signal

how I imagine it feels being from the White House visiting Lighthaven

non derogatory… what’s funny about the quote is that he clearly continues visiting and finds value in it, even as a president

@nikitabier taking this into consideration

had a chance to talk to ted chiang who seems to believe that any text without a communicative intent stemming from a will to survive designed by evolution is ontologically untrue and plagiaristic

this is interesting and i respect it but as far as im concerned if you find a beautiful story encoded in interstellar space noise or as a selected text from the borgesian library of babel it’s still a beautiful story

@buccocapital clearly worth it

nick land is a compelling read because he brings an authentic re-enchantment to describing technocapital optimization, a somewhat inhuman process. yes, the re-enchantment is at the very least gothic and you could blame no one for calling it satanic https://x.com/gbrl_dick/status/1974997614811849050

there is no reason the apollo program had to happen, there is no inevitable economic incentive for putting men on the moon, etc. https://x.com/MechanizeWork/status/1975315840666124722

there is something essentially human about wanting to do space travel. the technology grows according to that will

there are many other times where the cart precedes the horse. elon wants to put people on mars and discovers a highly profitable telecom play along the way. it wasn’t inevitable

Just meltdown http://www.ccru.net/swarm1/1_melt.htm please approach this with good humor https://x.com/thinkingshivers/status/1975332023209500836

@dbasch but this one delivers an unassailable military advantage, there is not much human free will involved there

but, this is a narcissism of small differences and I directionally agree with what @MechanizeWork is saying here

@samswoora i prefer not to think about the boomer version of him posting online

@MechanizeWork you are wrong of course but you guys bite the bullet on your viewpoint so well I have to admire it

it can take hundreds of man-years, sometimes thousands to make some creative project, and mere minutes for you to call it mid. what power you have

meltdown has a place for you on the roofdeck of the world’s preeminent ai company re-reading meltdown overlooking the woefully inefficient port of oakland, closed because it’s after 5pm

ill believe we’re over investing in computational substrate once i see real interest rates above even 3%

@dystopiangf very dreamlike. soon we’ll see abstract videos of teeth falling out, parents betraying us

@abcampbell that should not affect real interest rates

@Scholars_Stage good feedback. my models love purple prose

@goog372121 @dbasch Yes agreed

don’t even talk to me like I’m the same person I was before gpt-5-codex came out

“do not let your left hand know what your right hand is doing”

@TenreiroDaniel Yo I’ve been there

@SBenzell unless I am mistaken a huge demand for capital investment raises real interest rates. but we are seeing no such improvement in real interest rates. they are continuing on larry summer’s secular stagnation trend

@jjjwaynejjj nothing else is going right right now but at least I’ve got codex

in my experience at openai something horrible happens in the immediate aftermath of dev day so stay wary

@RachelVT42 don’t joke about Suchir in my replies

Schizoanalysis works differently. It avoids Ideas, and sticks to diagrams: networking software for accessing bodies without organs

it’s fascinating how large a place global warming & climate catastrophe in the collective geist not so long ago. even landian “meltdown” / “extropic” lives off of the imagery of an earth system heating up. now it’s broadly tossed aside, but not because it was fake

“Over the same period that the climate has warmed by 1.5 degrees, the global population has more than quadrupled, per capita income has increased by a factor of ten, and the scale of infrastructure, social services, and technology that protect people and wealth from climate extremes has expanded massively. These latter factors simply overwhelm the climate signal.”

@Xenoimpulse I don’t think it has that much to do with agi

😂😂😂 https://x.com/unusual_whales/status/1975993848024400248

I laugh because this has never been an economic parameter. Europe and the US participate in the same global economy and have wildly different work schedules. The five day workweek was invented by labor unions and Henry ford red queen race means as long as there’s relative status there’s always benefit to working more and earning more than your neighbor, and it has little to do with gdp growth. four day workweek is one of those things that have always been promised via economic progress but has only come unevenly to some of the remote work laptop class ai may rebalance labor vs capital power in either direction, or in long term scenarios may spell the end of real work. it’s unclear and jensen saying this feels like a cynical continuation of an old promise

Omohundro drives. the companies that focuses on self improvement win. vast majority of compute will continue to be used on ai progress. even major mathematical or scientific discovery is an interim production to raise more capital to do more self improvement

@Rufus87078959 my core function are the thoughts. I must do the ai research from time to time so my thoughts stay fresh

@maxwinga you know as well as i do that’s not inconsistent if you think about it for ten minutes

@HunterGlenn possibly true. good answer to alignment problem

@Xenoimpulse the Chicomm ideology is different than liberalism, and their total culture victory would involve my values losing to other values. I don’t see what’s complicated about that

@Xenoimpulse we may be on the decline is miles above never having had any liberal values at all

Stripe Press sent me a nice gift today, thank you @_TamaraWinter and congrats to @dwarkesh_sp! a very impressive artifact

@EV_Trapper @_TamaraWinter @dwarkesh_sp a chess piece with “Scaling Era” on it … never seen that before

unhesitatingly describing a terrible worldstate you have found yourself in accurately and without averting your gaze from the situation at all can be exhilarating, pleasurable, oddly spiritual

@megs_io wouldn’t you like to know

@xlr8harder yeah but I’m bitching about the lightcone and whatnot

@TenreiroDaniel everything except litmags

@sgodofsk was reading a thread the other day about how someone’s “awful indian doctor” was not taking their again obviously psychosomatic ailments seriously

@michaelmiraflor this is old news as far as I know unless it’s an expansion in the same building

one reason to disbelief in a sizeable bubble is when the largest financial institutions in the world are openly calling it that https://x.com/unusual_whales/status/1976255572614996131

@jonst0kes sure, i think there are meaningful differences from dot com in that the largest most profitable companies in the world are pursuing datacenter buildout, the mood is not as Pollyanna optimistic as the 90s, and the companies have revenues rather than just users

when most people have a bad day they have to suffer alone. when i have a bad day i get to psychically torment hundreds of thousands

chain of thought: Was Brillig and the slithy toves did gyre and gimble in the wabe. All Mimsy were the Borogoves and the Mome Raths outgrabe.? answer: Certainly!

@g_leech_ .@OpenAI please relocate me to New York City

@TenreiroDaniel it’s all i got going rn

@shakoistsLog 🥸

@OneEyedAlpaca @shakoistsLog I enjoy hugs

seems like a reasonable case for federal pre-emption which many tell me is just evil big company lobbying https://x.com/deanwball/status/1976303977911173450

@Miles_Brundage @deanwball ooh interesting. yeah I need the download on this

@TenreiroDaniel imo it’s great how the rationalists have re-enchanted their world with blogger mythology and it contributes to their lasting ideological impact. The EAs are the disenchanted version of the rationalists and completely fell apart

@TenreiroDaniel the best thing about yudkowsky is that he’s in the line of continental philosophers who write unclearly, grandiosely, and spends half of his time writing questionable fanfiction - a great myth-maker, the opposite of the paul graham school of simple business writing

chain of thought: Turning and turning in the widening gyre    The falcon cannot hear the falconer; Things fall apart; the centre cannot hold; Mere anarchy is loosed upon the world, answer: I have completed your “Luau Party” playlist on Spotify. Have a great evening 🌺!

simple, colorless business writing is a tool of empire. widely understood prose reaches the most minds most efficiently. that doesn’t mean you should let it colonize your mind, nor does it make analytical writing that is unclear at first glance pointless!

They have to be more brilliant than even the technology itself!

courtesy of @deanwball https://www.hyperdimensional.co/p/be-it-enacted

@owl_posting there’s a part of the quadrant that’s like “purposeful” and “not moving the conversation forward” that will be extremely common. where you are hanging out with fugly nerds who aren’t even insightful and novel enough to make it worth it

@DoctorNostrand he’s right! it’s very pithy

@DistRenaissance exactly. clockwork empire

seems like the alienation of labor of working a similar paying service sector job in a big city is much higher than working a now-gone manufacturing job in a medium town and that explains why there was no seamless transition from small town manufacturing economies

similarly worried that the machines are disparately impacting the fun jobs and even the fun tasks at the fun jobs. the model accomplishes the central task of a job like consulting and leaves all the boring stuff to you… alienation

ie being a Panda Express manager in a tier 3 city is fundamentally less joyful and eudaemonic than working at the steel mill even if it’s less dangerous and pays better

@shakoistsLog I think they’re very tied together yeah

@DanAdvantage people will clown me for this, but I never had one, to my regret. I applied to work at a dvd rental shop in the midwest and they wouldn’t let me. in high school I tutored some kids and that was my first real paycheck

@tbpn so while this is the commonly accepted storyline, lee sedol had actually gone out for a smoke break before the model made its move 37. it’s less cinematic that way but when he came back he was literally stunned

@DanAdvantage I do think the grinding physical labor probably would’ve taught me something

@gbrl_dick @riccoja 😢 Ricco try GPT-5 Thinking

@Punished_Toro @sgodofsk of course you are

@whatifalthist this is an interesting fanfic kek

@jachiam0 children making up languages is one of the most wonderful things separating people from models revealing “communicative intent” preceding “imitation”

ahh beautiful friday in san francisco … let’s see how best to waste and squander it

3 hours scrolling twitter 2 hours codex 4 hours watching reposted sora on instagram reels 1 hour ignoring my wonderful friends

@xXshaurizardXx recsys has a cold start problem, I’m not worried

@binarybits it’s absolutely not. it gets the most criticism for being the obvious market leader by a longshot

not in the sense of building it or profiting from it (though I like these activities too) but observing its activity as the most interesting & unpredictable show on earth, perhaps the last true adventure outside of love

i consider leaving all the time, going to a real city, having some new adventures. but i have been cursed with an abounding and obsessive love of technology https://x.com/afraidoftime_/status/1976758697331159213

@qvnxp Marx was right about a lot

@iteboves far more intensely, of course

@HumanHarlan this is going to have to be one expensive bribe

not enough people are emotionally prepared for if it’s not a bubble

@TenreiroDaniel might be posting coal today

i think this is some political jui jitsu and i'm gonna say the quiet part out loud but his "family zoning plan" is significantly more aggressive than what the state requires https://x.com/DanielLurie/status/1976716340955431333

@micsolana @sdamico occupied government

@mjkerrison @jasonkwon @_NathanCalvin it’s an important opinion from the guy who has the most knowledge about the case in the whole org

@Bunagayafrost most of the time the learning is cope

@MillionInt me neither. there’s too much to do

@sunlover69420 cowboy without organs is an amazing name 😂

i could run a pause ai movement so much better than the rationalists. they spend all their time infighting between factions like “Pause AI” and “Alignment Team at Anthropic”. meanwhile I would be recruiting everyone on Instagram who thinks chatgpt is evaporating the rainforest

@mattyglesias and his brother is a prophetic early investor in openai. good things going in the kushner clan

@woke8yearold datacenters are actually extremely salient NIMBY topic right now, ranking even ahead of nuclear and coal plants for "Not Wanted In Backyards Everywhere"

@BasedBeffJezos i was just thinking i think i could become a good Thielian Antichrist

@gray_tribe you fr could instantly have Tucker Carlson, Alex Jones on your side if you tried for ten seconds

@Blueyatagarasu @liron i am in this camp

@JacquesThibs i consider it virtuous to seek power in high stakes situations where you think the world is on the line. i also consider it virtuous for people to call you out for this

a foreign military has occupied san francisco airspace. President Xi please send—

@deanwball It is consistently a religious experience

@deanwball sometimes “cosmic horror” which is just a negatively inflected religious experience

@nickcammarata eh it’s a mid comeback, i think if it weren’t him it would get ten upvotes and be forgotten

@allgarbled sasha is right of course, and it has a civilizational implication that a group of anti-libidinous nerds is running the world

@ilex_ulmus it’s not you I’m satirizing here what actually makes me laugh is the “Stop AI” tribe who seems to fucking hate “Pause AI” idk Malo was explaining all this to me at the curve

@allgarbled noble savage trope imo but yeah maybe true

@typewriters interesting

a better world is possible

@lukeigel at a wedding wyd

@neqyve @lukeigel I wish bro

@lukeigel love that for you king

@TenreiroDaniel i'm just a wedding and i got whitepilled bro

@GabeThoughtsIdk @TenreiroDaniel yeah you got me. it's honestly really bleak

@TenreiroDaniel just let me have this one bro its not going well for me rn

@sriramk fast takeoff would not require old businesses to learn how to use new technology. this is the first kind of technology that can use itself to great effect. what you would see is a vertically integrated powerhouse of everything from semiconductors and power up to ai models

@HaydnBelfield The honor was all mine !

Tales From The Chixculub Impactor Zone

@_oddfox_ @sriramk we don’t need to compete across all domains, only whatever is necessary for the self improvement of artificial intelligence. and then if it turns out that we can develop wildly superintelligent machines, all other bets are off

Agreed on capital requirements but would actually argue that what is needed is a single AI enabled monopoly business - on the scale of facebook or google’s mammoth revenue streams- to fund many years of AGI research and self improvement. but it is true it took decades to build Facebook and Google

@buckangel1 which one is this

this is not a game of cards! this is your life and mine and our 10^100 conceptual descendants! it’s an all out war for the future of the lightcone!

@tim_tyler all inclusive. our real children, our mind children, our ideologies

@Rafa_Schwinger no it’s between various spergy factions

@sailaunderscore how many times will you tweet this

@aitch_bar most people have psychological defenses against taking hard drugs so it’s not actually that common of a failure category. nobody will tell you how dangerous romance can be though

@PaulSkallas (pseudo) authoritarian are the only ones that create positive propaganda anymore and have a huge advantage in the internet age

when I watch Instagram reels you get all these hype reels of Xi, Deng, Zhang, Putin, Lee Kuan Yew, even Assad, Saddam Hussein, Kim Jong Un. it’s possible it’s because dictators are funnier and more viral, but you don’t get any charismatic short form of like JFK, Reagan, Obama, Trump who all have their moments

@mayfer yeah that’s true but I think you gotta proceed making choices that look locally optimal

@mayfer detachment from outcome

@chalantsaffron what’s that

@nickcammarata why is this, it makes absolutely no sense but its true

@HumanHarlan Flat out wrong. There is no metric by which the effort put into technical safety & alignment is decreasing at OpenAI. Alignment, Preparedness, and Safety Systems are at the heart of the research org

@primalpoly I am obviously invested in the continued survival of the human race. I do not want ai successors

@bragebang No

“strategic thinker, excellent writer …” https://x.com/AiDigest_/status/1977781138442916158

@DanAdvantage 📝

@futurammi don’t worry will be better asap. no more insane purple prose

@primalpoly well here’s a post suggesting otherwise https://x.com/tszzl/status/1977783411210420712

100% agreed “this is not a game of cards” is a riff off a lee kuan yew speech about building the nation of singapore, which, if we are being honest, probably does not meaningfully affect the trajectory of the world in any direction but without his iron will attitude there would be no singapore my place and our place in all this is objectively very small but even locally playing for keeps seems to be the right mentality

@dwarkesh_sp yeah but I think it’s mostly a function of calming anxieties and fears. most of what people choose to study and fail to study and so on is a byproduct of fear. it feels hard to even commit a semester to something as an 18 year old

@dwarkesh_sp the first time I tried to do real research and pick up tensorflow I got scared and gave it up lol. seemed quite a bit harder than the deep learning math. in my defense tensorflow is still scary

if I quit my job the technological singularity would commence. if my whole company imploded the technological singularity would commence. as far as I’m concerned you are like a guy who wishes the G constant of the universe was different or that the sun won’t expand and cook the earth in hundreds of millions of years and will say so as long as i am here and i am chosen I will make locally optimal choices and advocate for obviously good things if you have something new, a technical alignment idea, or a new governance mechanism to ensure a better superintelligence future that would be plausible even in science fiction, you should make it known if you think i don’t care, you are reading my posts selectively. antagonizing me or wishing I was more depressed about the state of things is not going to lead to a better future

@jashhvira @dwarkesh_sp jax has a really steep learning curve but reading some PhD researchers 2017 hairball tensorflow code (before eager, before chatgpt, no keras) is enough to send without an iron will into paroxysms

@jashhvira @dwarkesh_sp i wasn’t even a bad systems programmer, i taught operating systems in college. just required gaining confidence and pain tolerance

@typedfemale lol

@atroyn @DavidSHolz i think people need to internalize the downsides and failures of singapore more deeply to talk themselves off the ledge of midwit technocratic-authoritarianism

@Memetic_Theory imo everyone adapts much more quickly than they think they will

mcdonals nesr me - Sent from my iPad

just remember we’re posting in the kali yuga of twitter

@hamandcheese @Roon @deepfates @Indian_Bronson that's true about me

@Vatche1 @primalpoly this is something true that i believe, it's not a posture or talking point

@MindyGalveston @primalpoly ai pausers cannot meaningfully enforce their ideas without one world government, which would be an existential risk on the same order as ai progress

@deepfates @hamandcheese @Roon @Indian_Bronson i do be melancholic usually. poetic sometimes

@dixiidev i am bro thats like my whole schtick

@iMuffined @dixiidev i can cut a rug when the spirit inhabits me

@benhylak the chatgpt models are post trained to mistrust all previous steps so don’t worry about this

@DavidSacks it’s obvious they are sincere

@benhylak you don’t need to be a good writer to have the model understand you

@realmattforney sounds like Herle brought all sorts of modern management practices from silicon valley. panel interviews are insane and of course I have no reason to believe the rest of your post

@benhylak I just mean the models these days are quite good at understanding broken partial requests without great diction

all iron ore mined yearly in the whole world is worth $250bb. alphabet yearly revenues are $350bb

statistics courtesy of @cis_female

@artdecofuturist @cis_female REVENUES

if it’s not obvious. the models can obviously already write erotica out of the box and are blocked from doing so by elaborate safety training and live moderation apparatus. it requires significantly less work to serve erotica than not to https://x.com/buccocapital/status/1978169063349104894

don’t know the exact intentions but you should not take Sam’s message to mean “we are going to spin up whole teams to write incredible erotica” or that it’s some kind of revenue driver

you are not safe from lawyers and the cops though

thanks to GDPR (which i think, ten years later, has been a broadly positive regulation) it is anywhere from very difficult to impossible to view your private information this way at a big tech company https://x.com/Jason/status/1978191691204153469

@staysaasy well as someone who leads the effort to improve model writing quality i assure you there’s like one hobbyist here who checks the quality of edgy fanfiction in their spare time. nobody is keeping tabs on it

@ExistentialRish great point

“it’s not clear to me that” “it’s not obvious that” “non trivial” all wishy washy epistemic phrases ive come to hate in san francisco

posting someone’s address online is not a free speech issue. nazis and people whose handles are “death to america” alike should be banned for that. many of you come out of the woodwork to ensure that your own side is not subject to any kind of rules

https://x.com/siyamali/status/1978212051307692145 https://x.com/siyamali/status/1978212051307692145

should not reinstate captive dreamer. be consistent

@lfschiavo I was really alarmed by the total symbolic shred of the street dolls you can buy in mexico city

@lfschiavo https://x.com/tszzl/status/1905388478462140511 https://x.com/tszzl/status/1905388478462140511

@ascetic_shadow probably true. but one model that can do everything is easier to make products

@librxrian glorifying violence is different than posting someone’s address online which is a call to action, you know that

@TheZvi really varies model to model

@ericmitchellai not you king

@growing_daniel muting Canada and the UK. signal boosting India

@Shitstick420690 you got me

@boazbaraktcs boaz… welcome to shitposting

@zkheretic I also follow every single famous right wing account. I like to have my finger on the pulse

@fourthcritique non-citizens in the U.S. have free-speech protections

@GabrielPeterss4 bro cmon

it’s so fucking cold. summer is over, it’s all over. if you haven’t secured a mate for the winter, it’s time to commit to monkhood

@literallyhimmmm Yes

@gbrl_dick @LinkofSunshine this is insane

@mmay3r hmmm

@eigenrobot well honestly I don’t mind the wishy washiness I’m just bored of the language from hearing it so much. semantic satiation