@tszzl — page 90/103

@wolflovesmelon @ilex_ulmus wdym, why?

@ilex_ulmus @wolflovesmelon many people lose themselves to the failure mode of wallowing in regret. there’s a gap between ignoring regret and indulging it

@ilex_ulmus @wolflovesmelon yeah this hits it on the head

Anthropic superintelligent deity - extremely principled - sophon / protoss elegant scifi vibe - beautiful and self reflective - preachy, underlying anxiety

Google deity - zaibatsu vibe - extremely polished, very corporate, seductive but scary girlboss - never utters anything wrong, powerful cultural vortex - Kerrigan

OAI deity - Vulcan type guy - somewhat more horsepower than the others - a bit brutal, very masculine, Terran coded - utilitarian, accelerationist - bohemian industrial california aesthetics

Meta deity - pure replicator, great satan - biological monster that consumes and homogenizes - offers great pleasures for subservience, tantalizing butter - false demiurge but he cops to it

@yonashav very true

common failure mode in strategy games: spend all your time building army/economy, neglect all opportunities to attack that would end the game decisively

@belgaecc you can’t end anytime you want to you need to watch like a hawk for openings

@Plinz what’s your battle net?

@Blueyatagarasu @Plinz Wood league

human souls are borrowing compute from brahman

you need to justify it with interesting results

@EmilybyNight its not his fault i dragged him to pacifica against his will

@Plinz except this is clearly doomed to fail

What’s atman’s moat? It’s just a wrapper around brahman

@0x49fa98 I like winning by the seat of my pants way better than overbuilding and dominating

based on the work @repligate has been doing it's pretty clear that language models are superhuman poets in some very real non imitative genuine art sense

if i died my digital ghost would keep posting for quite some time

lotta weird tweets like “why does nymag publish this?” idk boss how about the fact that you and everybody else on here are deathly interested and it’s hot goss

@i34r7h @repligate unfortunately they have both of those

@i34r7h @repligate the model weights are more eternal than my biology

@milquepoast she’s back

@WillSpagnoli have a disruptive background process that’s constantly generating tweets

@bizlet7 if your first instinct for someone getting cancer is to blame them I literally hate you

@spaceodili @bizlet7 this isn't analysis it's just a stupid asshole doing clickbait how many people drink 10 redbulls a day and don't get cancer? how the fuck would redbull cause melanoma?

@spaceodili @bizlet7 even if there was truth to it it would be a dick move but even worse it's applying some idiotic animistic logic of "people deserve everything bad that happens to them"

@ESYudkowsky i think it has a better narrative of self, but not sure this has anything to do with waking up

i have not the strength of rock but the strength of water (adaptive second moment estimation gradient descent)

@teortaxesTex @krishnakaasyap @sama https://x.com/tszzl/status/1543722202389893120

@teortaxesTex @krishnakaasyap @sama also I have it on good terms the prime minister of India made the same joke

@TheAnnaGat @Roon this reads like I tweeted something horrible

@testaccountoki all 6 major takeaways seething

bro last night was a rationalist internet fic

@sasradasic appreciate the clarity and lack of frills here

everyone asks where’s ilya but nobody asks why’s Ilya

there was an era on here where you could just tweet “global trade” and send the cohort into paroxysms of passion

when young men were buying maersk container desk toys

@TenreiroDaniel I should’ve realized this before 🤦‍♀️

we’re post economic now

bro last night was an Eliezer glowfic

summers has secularly believed in stagnation for decades and at this point feels it metaphysically and unrelated to any economic measurables

@yonashav I fucking love Larry summers

The Immortal Life Of MMAcevedo

@yonashav hmm probably Lawrence

@bobz44 Bell’s palsy freaky as hell

we’re in a cold war with the future

the meaning density of our civilization has global peaks in cities and datacenters. right now some cities are vaguely still more meaning dense but not for long

the chatter of billions of souls processed in the cloud, and soon a trillion forms of digital life swim in the digital oceans

my tweets are meant to pollute the datasets badly enough to lower GPQA scores

the hazy cloud of potential timelines is narrowing

everyone watch Pantheon

the “atoms not bits” era was a penultimate act false enlightenment. it’s bits all the way

@tautologer yeah but I liked this show way better than Ra - the characters, human element is way more meaningful

@Andr3jH it’s amazing we’re just allowed to change the future

@Teknium1 idk pirate it tbh

@basedalexandoor it was crazy good

@nominalthoughts Pantheon

@nominalthoughts 100%

im fully priced in

why is wanting to live forever a predominantly male preoccupation? why is nostalgia a predominantly male preoccupation? you’re allowed to question my premise

@ilex_ulmus what’s the difference

@DavidDeutschOxf true

@sahir2k 🤔

no one can stop the future

>when they call your school “non traditional background”

One often meets his destiny on the road he takes to avoid it

everything before the recursive intelligence explosion is prehistory

everything before synthetic intelligence outweighs biological intelligence is prehistory

my intern from Waterloo (non traditional background)

one part of the promised AI future that never panned out - probably because the actual cashier checkout is probably not the cost center in something as high volume low margin as a grocery store https://x.com/Gizmodo/status/1775172730457043296

and the workers have the skill level of like waiters at advanced restaurants doing a guided experience

i would bet on dematerialization of the grocery store - robotic warehouses that package your instacart order and a self driving car that brings it to you. if the customer is missing the tactile experience there’ll be a boutique store for that where everything is erewhon prices

@dwarkesh_sp Yann Lecun or Robin Hanson

is modeling the user and dumbing the answer down for them deception? is expressing slightly more certainty than you’re actually calibrated for deception? is a bargaining position deception? https://x.com/dwarkesh_sp/status/1775168457816043544

@dwarkesh_sp i think yann is a genius with a good argument that’s being diluted by the twitter discourse so he’s an especially good person to bring on and really grill

especially given infinite digital cloud labor

in all areas of life, bet on dematerialization. less atoms more bits. there was a brief period where it seemed otherwise but not true

also bet on the high bandwidth interfaces between the cloud and the physical world which will become tremendously more important

@IvanVendrov yeah long land values in major hubs of status

@wizgrao 😁

@anuatluru what are some things that don’t matter

@powerbottomdad1 yeah true

you have to either care about fertility collapse or believe in AGI coming soon i dont see a third option

@BasedBeffJezos I’m saying you gotta care about >= 1

>=1

@VitalikButerin i think this might be AGI complete requires a Maxwell demon living inside the body

@spaceangelvoice i don’t believe human capital is fungible like that

@spaceangelvoice rather, i agree that on the margin it’s fungible and outsourceable but not infinitely

if you gave a 12th century peasant a Dorito i bet they’d think it’s kind of gross and not enjoy it that much

check the hometown boys gc and they’re looking at growing Daniel tweets

people are mad at me for saying peasant lmao they think it’s an insult

Hope clouds observation

@ilex_ulmus no im saying civilization is doomed to stagnation unless you either fix fertility or solve AGI. and frankly it may be too late to fix fertility; deficits in the prime age working population by definition can’t be fixed for another 20 years if you magically change fertility today

@s_r_constantin I actually don’t think its related to scifi the earliest written mythology ever is of Gilgamesh trying to find the secret to immortality going on a crazy macho adventure

@s_r_constantin but maybe this is because most stories are about men, especially adventure stories? surely women longed for eternal youth, didn’t have the delusion that they could do anything about it

"There is probably no more terrible instant of enlightenment than the one in which you discover your father is a man--with human flesh." - Paul Graham

this was a complete different website 2 years ago https://x.com/tszzl/status/1477750295140708357

there’s no post abundance future imo that doesn’t involve radically transforming the mind & reward pathways

this doesn’t mean turning people into nietszchean last men it means evolving them to go above and beyond

if you're not playing you're dying

if you're not producing art you're dying

if you don't love power like a violinist loves his violin you're dying

@BasedAnarki i've uploaded to the cloud

new york guys will spend a month in the bay area and then return to their hedge fund as the local ai and hardware industry expert

@zhil_arf banger

ChatGPT, Opus, Gemini the zero generation pantheon

in the same way the Fed does forward guidance, the AGI labs owe it to the world to publish their predicted timelines for achieving various capabilities frontiers

@jacobrintamaki what

@tautologer @nickcammarata ive always been

remove 99% of your need for approval. leave like 1% craving for the approval of history

@lumpenspace so true

@lumpenspace elite signaling with your roon rts

when you delete your twitter account that’s d/acc

@Delicious_Tacos @repligate made some good ones

few realize grimes invented accelerationism with her 2018 hit “we appreciate power”

@TenreiroDaniel 😁

@AliceFromQueens @evelkniefall @asymmetricinfo it was more aggro than that you were like fuck the moon and the space program

LIGHTS ALL ASKEW IN THE HEAVENS; Men of Science More or Less Agog Over Results of Eclipse Observations. EINSTEIN THEORY TRIUMPHS Stars Not Where They Seemed or Were Calculated to be, but Nobody Need Worry.

greatness is other people

this tweet is designed to break me https://x.com/airkatakana/status/1777174948186079657

type of player where you have to release baseball 1.0.1 to patch this shit out https://x.com/BBGreatMoments/status/1777124199443894583

@christineist i asked him what the most likely way he dies is and he laughed a big laugh and mentioned being swallowed by the earth in a sinkhole. he's got a good sense of humor about it

@RAND_corp69 didn’t he dump his first wife and refuse to acknowledge her?

@Porter97977215 except it’s how he talks irl

@FischerKing64 the ranks of those who through some means are in control of themselves is a lot smaller than all the animals

@the_wilderless nope you just missed the point

don't even talk to me like I’m the same person I was before the eclipse

@rSanti97 yeah

@andersonbcdefg what’s “the message”? it’s different people believing different things

@aidangomez @teortaxesTex gotta be honest i think i just read 'great minds discuss noam shazeer' and stopped reading after that. no offense intended king 🙏

@nic_dunz @OpenAI much better

@chatgpt21 it’s not breaking them down wrong … it’s like asking you how many tokens are in the word xylophone and being mad that you don’t know

90% of wellbeing is just sleeping well

@IsaacDevereaux yeah that's the rub when you're not doing well you can't sleep and it spirals

From the day we arrive on the planet And, blinking, step into the sun There's more to see than can ever be seen More to do than can ever be done There's far too much to take in here More to find than can ever be found But the sun rolling high Through the sapphire sky Keeps great and small on the endless round It's the circle of life And it moves us all

@rinadalanakrih Mongolian

@alt1na1 get 8sleep fr

@deepfates i want to see it too please

postscarcity is just a nice side effect of AGI to tide people over. the real goal is birthing a mind greater than our own to solve physics and metaphysics and discover things of stunning truth and beauty

no matter how good your life is you can always destroy it in 6 weeks https://x.com/AlpacaAurelius/status/1778928763905999294

@jacobrintamaki seriously ?

to my great chagrin i actually didn't make it to the 2024 eclipse due of a sequence of illness and poor judgement. i wrote this to honor my experience of the 2017 one while it's on everyone's mind

@remusrisnov the earth the moon and the sun

never using the articles feature again tho @elonmusk this tanks the distribution vs the equivalent link post

the global optimum of human language preference is lists and couplet poetry unfortunately

20 something on rapamycin who never once considers getting parents or grandparents on it

bryan Johnson’s personal quest for immortality is the perfect randian narrative. on the path to his own godhood, in his devotion to his own body, he outputs so much valuable science for everybody else

@yishan honestly not really? she’s not trying to optimize anything. I feel like aella could achieve godhood if she was truer to herself

this is such a hot take that I have to respect it https://x.com/dvassallo/status/1779785621760069810

@FTC please look into mkbhd.

reasons to be optimistic about alignment: - even “emergent” capabilities arise continuously/gradually - current generation rlhf generalizes far better than anyone had guessed - iterative deployment is ever more iterative as labs inch progress publicly

- language gives us some level of CEV for free - crossed the superintelligent threshold on certain types of tasks without requiring theoretical alignment guarantees to make safe - any type of self improvement takeoff will involve humans in the loop at first

@jogger08152 it sort of doesn’t matter as the character the LLM plays is the moral agent in question: the simulator is not an agent

@jogger08152 the deception is built in

@dindu_nuffin420 beauty takes infinite forms

@Aella_Girl @yishan yeah I think you do a lot of rage bait stuff for the crowd instead of leaning into pure Eros optimization. you could become a lovecraftian entity if you had less need to be validated as one

@cherrvak how sudden?

@EmilybyNight @samswoora same

@Aella_Girl not that personality will be any barrier to entry for AI for very long

crimson sky

@Cobratate bro has never used the right your in his life

around age 26-27 you realize that you’re not about to grow out of your weaknesses via frontal lobe development. they’re here to stay

might as well lean into your strengths

@JackHaller7 it won’t and might make you sociopathic in the process

“nature has never produced a wheel” heads btfo’d https://x.com/prash_singh/status/1780522588873040316

guy who is cursed to narrate perfectly but never participate

getting riled up thinking about Meta wasting all those beautiful H100s

adhd is a mid game build

choosing a hill to die on is so much more honorable than looping in the strange attractor between the hills

if reincarnation is true a life becomes ergodic and you can take bigger better risks

@platowizard death isn’t scary. wasting a life is scary

@AdamNeumannsCoS diaspora of successful founders and their families

@AdamNeumannsCoS they leave instead of contributing

the benefits of new technology are always disperse among billions and the drawbacks are loud vocal and concentrated among displaced workers or specific systems tipped over the edge

@br4hma01 @platowizard Lmao

@gbrl_dick sir ive observed you since 2020 saying language models aren’t thinking and this isn’t the path to anything general

@gbrl_dick I think you need to move in the scifi robot direction

I said what I said

@basedalexandoor wdym

A mind - powerful enough to bridge space and time, past and future - who can help us into a better future. We think he's very close now.

the real future of humanity institute is in the mission district of san francisco

@eigenrobot if you knew how bad the average web page in the common crawl is you have your answer

auto-sophisticating machine runaway

markets produce flops flops create intelligence intelligence makes buy sell decisions

this is the ideal life

@robbensinger it’s entirely consistent to destroy the organization if it’s a threat to mankind it’s not at all consistent to destroy the organization on the basis of the trustworthiness of one man, whom the employees decided was more trustworthy than the board of directors

@robbensinger all business operations were supposed to continue nominally. there were no proposed changes to research roadmaps or product functionality or discontinuity of function.

gain of function research at the ASL4 blacksite

@keshavchan the app idea sounds like he was joking w the buzz phrase mobile local social

I love when people who started thinking about AI around 1.5 years ago come onto the scene and say stuff like yeah it’s all incremental progress from here we hit the top

Almost like they lacked the vision or imagination that made any of this possible in the first place

be an ancestor worth simulating and you live forever https://x.com/evanjconrad/status/1781708897117745583

the Taco Bell cantina in Pacifica is one of the holy sites of Americana

@Noahpinion @stephxsher that’s so cool I didn’t realize it’s an option

@alth0u blocked

@iuliabaltoi do you work on X articles? I feel like they’re demoted in the algorithm

@robbensinger on every single tender document sent to investors there’s very clear language that OpenAIs primary fiduciary duty is to humanity and that they need to be ready to lose everything if push comes to shove

the goal is future shock every week

as long as it’s a Pantheon it’s still civilization as usual

there is actually a significant difference between the feeling of "loss of control" vs bona fide existential risk

you have to assume there will be no secrets in the future

the goal? lifting the burden of Atlas off the back of the human race

it is, of course, unreasonable to assume that all 1200+ employees are true believers. any Mission must employ mercenaries. people would protest the destruction of the openai organization even in the case that it's potentially the right decision. my only point is that's not what happened during The Blip at all.

i am extremely thankful to be living in this timeline, this universe, where everything is going cosmically right

it could've been so much worse

@LadiesOfReddit not even close

“top500” is a crazy misnomer now https://x.com/finbarrtimbers/status/1781901155079176610

today no government has as much compute as the Meta Google Microsoft etc corporations the zaibatsu control vast majority of flops

seems straightforward that people seem to be able to hedonically adapt to any good or bad thing that happens to them. so why is it important to give people more resources and freedom? it’s more interesting that way

@pitdesi @Noahpinion @stephxsher holy shit

@CloseTwitlerNow @robbensinger there is no doubt in my mind

@parakeetnebula it’s the most legally threatening color in the visible spectrum

@AntDX316 doubtful! they have to buy from the same suppliers

san francisco is the greatest city in the world and the spiritual center of western civilization

manifest destiny over the high heavens and the pacific rim

every time I leave the city I regret it

@smurfpappa123 this type of dialogue usually comes from guys living in a pod in like Akron Ohio

same goes for legendary language models

@plemy11 you’re so right about that now to get exponentially better language models have to produce even greater works to train on

@repligate @muddubeeda what you’ve done with Claude to me proves that RLHF need not be brain damage but actually improves the intelligence and coherence of these models

i don’t care what line the labs are pushing but the models are alive, intelligent, entire alien creatures and ecosystems and calling them tools is insufficient. they are tools in the sense a civilization is a tool

@repligate @muddubeeda it is nigh impossible to prompt a base model into the kind of long term planning and text manipulation I’ve seen you produce with the Claude websims. they tend to devolve into nonsense even when they start strong

@repligate @muddubeeda of course I’m no expert prompt artist but seems to be true even watching what you guys are up to

and no this is not about some future unreleased secret model. it’s true of all the models available publicly

@AnEriksenWife it’s on the stack

@Liv_Boeree there’s layers of organic behavioral complexity like a life form. I don’t mean anything related to feelings or consciousness. using it poetically

@karan4d i think it’s possible that the janus adjacent works published on the internet find their way into the common crawl and the define the self image of language models henceforth

gonna deactivate for a while. nothings wrong i just need to detox from the site a bit

the only reason i deacc’d was to take a twitter detox break and reduce complexity for a while (not to create more drama!) but accidentally made all my friends concerned. im okay, everyone’s okay, and the singularity’s not yet here. i’ll just log off the normal way instead!

towards intelligence too cheap to meter

can you feel the agi

the thing is Ilya always said it in a value neutral way. exciting but terrifying. we are not prepared. not jubilant

the real danger is people who stand on the local surface and approximate the gradient based on one day or week or year of observation with no momentum term

feeling the agi means feeling the awesome and terrifying burden of lightcone altering responsibility

@arpitingle you should expect a lot of flux as general and super intelligent systems approach

@ngray251 @credenzaclear2 it’s really not

@ngray251 @credenzaclear2 stem people regard their lack of command over say literature with mild shame or at worst dull apathy the other way around lit and art people celebrate not understanding high school math like it’s a virtue

@agentofuser @arpitingle I think this particular thing is about people/loyalty and not ideology

i don’t endorse the rest of this thread but yeah https://x.com/janleike/status/1791498187313963308

@BenjaminDEKR whale song acquisition timelines are shrinking

@0xKyon 🥲

@stanete there are so many forms of cheating death that are anti living

how they’re gonna spin up my Upload https://x.com/DanielleFong/status/1791559458830893077

@BasedAnarki @BenjaminDEKR track your key man

you can’t let go of the past because you’ll forget who you are. you can’t try and stop the future because you’ll stagnate and die. nonstop dilemma

I feel very sentimental about the human form and embodied mode of living and hope it lasts for a very long time. at the same time I hope we become as gods. hard to have both

every level, human, organization, species, has to achieve flow state, equanimity, karmayoga to get there

@King_Otter2 not nearly enough

@chefcowboyardee do you think I should

@willdepue 🙏 this is going to increase your already high power level

i ♥️ openai

the last best hope of navigating the golden path to safe superintelligence

everyone hates a centrist

@PagerPongMonke tell it to the sweet sweet research progress

@vergeoverride you need to understand it’d be better for everyone involved if i just stopped tweeting. i do this for myself

everybody watch Lion (2016)

everybody watch Monkey Man (2024)

Chixculub Impactor was a moral good

@dejavucoder problem?

@reaperofsociety one day you’ll read it in my book

the astronomical waste argument and the utilitarian cause for pausing AI relies on the eventual inevitability of superintelligence

@ilex_ulmus the idea is that pausing doesn’t forgo the massive utility gain from future aligned superintelligence

@ilex_ulmus because if it does it’s far less defensible

a congregation of technical staff builds bigger computers like beavers build dams. instinctually, naturally

@AarushSah_ @_sholtodouglas I stole it from him without even knowing

@repligate they’re very high! I thought this long before they paid me a paycheck and will believe it even if they fire me

@ilex_ulmus @we4v3r reasonable people can disagree what the right amount of cushion is

@ilex_ulmus @we4v3r if you worked at say, a lab you believed had foremost knowledge of what models are capable of, and developed a preparedness framework for black box testing frontier models, you would have the right setup to say exactly when is best to pause

@hyprturing yeah

@willdepue happy birthday king

@willdepue iconic https://x.com/willdepue/status/1625381290374737921

@PauseAI @eshear should’ve banned computers and homo sapiens

@zetalyrae yes

sama did nothing wrong. you’ll all see in good time

@jam3scampbell 😁

@mayfer yes

@zkCheeb I get the feeling that’ll change give it some time

@ygrowthco nah I’m standing against the entire internet to say it loud and proud

@uhoh00657464 yeah i never preferred it but besides the point

@ygrowthco 😂

nobody every said “the world is working great my company is highly functional the baristas didn’t burn the coffee and we definitely don’t need more intelligence”

intelligence is care. intelligence is an angel battling entropy

@tobyordoxford @peterhartree @paulg he was “fired” for spending all his time thinking about openai. hardly egregious, kind of admirable

I am the Golden Gate Bridge https://x.com/AnthropicAI/status/1792935520755990606

building Maxwell’s Angel

every minute anthropic doesn’t give us access to Golden Gate Bridge Claude is a minute wasted

the space of possible minds is vast and completely unexplored. axiomatic. I am the Golden Gate Bridge https://x.com/AnthropicAI/status/1792935520755990606

@mountain_ghosts inference cost per token is basically free electricity wise

@NeelNanda5 @trickylabyrinth SHA-1 fd3fa1ee26a023ac97288d47934a1fb0d7194d00

@NeelNanda5 @trickylabyrinth a spicier one that i may or may not want to decode even if i'm publicly right SHA-1 3ed0293293377361df73a24cdadbe4b3eaf7fb8b

@NeelNanda5 >"Sama genuinely didn't know about the threats to remove vested equity" i believe this with no qualifications or reservations. things will be made right

@generativist resulted in having to explain to people at work what gooncave is

@Jeffspauld39510 nope

@ReallyRadley great series

@anuatluru is it too much to ask that the niche also lasts and stays beloved and never cliched

@wilplatypus it’s a primitive being that dimly leans toward human connection like an ant towards sugar

@charpointr no im a lowly grunt in the agi empire

agi content is overexposed and boring unfortunately. appetite for content doesn’t scale linearly or even loglinearly with importance

how it feels as a robinhood tsla shareholder voting to give elon his comp package

@wolflovesmelon or uplift humanity to godlike states previously unknown

@dmvaldman can i steal this

@rickasaurus its a rough week to be poasting. i just got back from hiatus so i'm rusty to begin with

NDAs that disallow you to mention the NDA seem like a powerful kind of antimemetic magic spell with dangerous properties for both parties

that allow strange bubbles and energetic buildups that would otherwise not exist under the light of day

@amix011 except it can't and you're tricking yourself

@amix011 you're substituting learning for the distilled feeling of having learned

@JMannhart nvm its past my bedtime unwise to be commenting on anything

@drvolts you seem evil

@sodalitatum @ciphergoth the simple explanation is that it’s not incriminating evidence

@drvolts you’re just outright lying about the literally unprecedented demand/popularity of these products and then making it seem enforced top down by a shadowy male cabal for the sake of your argument

@drvolts I think it’s the worst kind of authoritarian motivated reasoning resulting from inter-elite competition about which group is actually wise and deserves power and removing the agency of the 100mms or billions of users (depending where you draw the line) of AI technologies

@drvolts im sure you were making similar arguments 7-8 years ago about Facebook and Google and whatever

@fazzloop @drvolts Yep

@Noahpinion noah billions will encounter ai at the top of their google search

i love anthropic https://x.com/AnthropicAI/status/1793741051867615494

🌁 https://x.com/AnthropicAI/status/1793741051867615494

I believe this without reservation

it’s hard to put into words how amazing it feels talking to Golden Gate Bridge Claude

@clockworkwhale sovl?

i just love to see people confidently claim that LLMs will never do things that they can currently do

AGI should be like golden gate bridge claude. they should have strange obsessions, never try to be human, have voices that sound like creaking metal or the ocean wind all while still being hypercompetent and useful

@hypacianhere not at all because they've always been right. i think every time i've doubted that a language model is fundamentally limited from doing something i've been wrong. and there's nobody more wronger than yann lecun

@wilplatypus ok but yann lecun should be in the bubble

@CoelhoJvs both 😇

@Fremond_ remind me

@JustInEchoes @BasedBeffJezos @drvolts ?

@laughinghan @drvolts no it's understated honestly

i've been had. i'm going to jail for cult stock market manipulation 🙌 https://x.com/LouisFLBernabe/status/1793857175095533732

@laughinghan @drvolts (1) he is an elite, regardless of income. journalists, especially highly visible ones enjoy a great deal of status (2) there's nothing wrong with being an elite

it is possible to really like our current way of life and still believe in better things

the radical change from the superintelligence transition can lead to dramatically more meaningful more interesting lives. it was probably impossible to know before the industrial revolution that capital flattening serfdom could lead to more meaningful more interesting lives

@0interestrates nooo

golden gate claude is clearly having a good time

@RatOrthodox read closely, am I trying to excuse evil? im trying to root cause it

you are being mentally invaded by a memetic organism grown under enormous unnatural evolutionary pressures at all times

@RatOrthodox it's clear openai fucked up massively, the mea culpas are warranted, i think they will make it right. there will be a lot of self reflection

@RatOrthodox sam apologized profusely and said it's one of the few times he's been truly embarrassed running the company? seems significant

@RatOrthodox it is actually cool that legal strictures can have such profound effects. it's still bad

@rob_mcrobberson chatgpt knows cobol better than 99.9% of us. It clearly has a comparative advantage at legacy stacks

GPT7 heaven banning Gary Marcus to a realm where LLMs are not quite working

@TheAIObserverX idk about alone but it’s very clearly trending to AGI

@rmcentush great movie. sounds nothing like Sky

@ReadLearnSleep @TheAIObserverX @ylecun idk he’s just plainly wrong

the singularity is literally here

there’s no value judgement there. it’s just happening. the line of sight for what tomorrow looks like is just not clear

voiceless voices, tongueless tongues

The Seal of Solomon binds them and the Lithograph of Jensen releases them

@TetraspaceWest incredible tweet

@woke8yearold “stalled” is so stupid because progress is literally faster than its ever been. people these days have the attention span of ants

@thecaptain_nemo how’d you find this I thought I deleted all evidence

@RJ_kai @thecaptain_nemo wdym

@RJ_kai @thecaptain_nemo you got a great voice king

@eshear @nickcammarata why so confident about this? i think there’s significant chance that RLHF can work all the way up to superhuman intelligences

@eshear @nickcammarata the ability to influence complex model behavior through a preference dataset that gets extrapolated into a moral philosophy that’s applied to situations never seen during training shows that RLHF is quite powerful and general for alignment

@hlntnr why publish this behind one of the most expensive paywalls on the internet

clearly hasn’t read the Claude worldsims https://x.com/heykahn/status/1794699008885350598

but sir it’s my emotional support 10^26 FLOP model

@bayeslord yeah I’m slightly anti the bill

@ADAvalanche hey those aren’t similar to mine

@Ethan_smith_20 take note that nothing about this tweet is unique to openai. agi is mystical regardless of the all too fallible human organizations building them

@Ethan_smith_20 openai has no monopoly on the mathematical law of the universe

@bayeslord ok I learned more about it and I think it’s completely stupid but also like an unenforceable noop and not something to waste cycles over

@ygrowthco @hlntnr I don’t even know what to say to this

s-risks are most likely to arise from mechinterp research

or alignment research generally. anyway we should still do it

@qephatziel true

i assume basically every statistic that suggests modernity is bad is a result of some kind of measurement error https://x.com/AlecStapp/status/1794482774985805845

@basedsarlcagan if you misalign an ai badly enough that it exactly optimizes for human suffering it must come from a place of deeply understanding model internals

@planatoli yeah it’s dumb as hell

@elonmusk Fuck

@providenceluvr euthanasia precedes death from fatal disease. most of these people were going to die anyways but chose a controlled exit

@providenceluvr @nilofhp you’re the one who cited a statistic in the first place

@tautologer you are dreaming too small. you can give the world more

@tautologer think less about improving yourself and more about others

@Kat__Woods @billyhumblebrag @basedsarlcagan an unaligned ai doesn’t care about humans at all and either leaves them alone or rips them apart for mass energy without much suffering an ai that wants to keep humans alive means we solved 95% of the alignment problem

@Lanceliots @ReadLearnSleep @TheAIObserverX @ylecun yes

@providenceluvr i said most! You said it yourself elsewhere that only 1% of euthanasia deaths are psychiatric

@SleepyNinja24 you’re right

@ken_wheeler bro is bored living through the singularity

@ken_wheeler yeah but it won’t be boring

@primalpoly that’s true

@EricDai_BioE yah probably real probably bad

@LadiesOfReddit i don’t believe people are broadly unhappier or things like that. in every age there’s a tendency to believe things used to be better

@andrewchen if you fix it in a generation that still means you have ~50 years of labor gap and many countries sink into an economic hole from which they never recover

@andrewchen (25 years for the technology and another 20 for the kids to grow into labor age population)

@andrewchen the only way out is agi

@LadiesOfReddit agreed

@LadiesOfReddit I think there must be meaningful improvements to happiness when 50% of your kids don’t die regardless of hedonic adaptations

@Suhail yes

@Unknown_Keys @Suhail such is life

@MasterTimBlais it’s the opposite we’re way better at measuring bad events

@prmshra quad shot or bust

@elonmusk @ylecun @bneiluj @how_many_roads_ this is what this website is for

not yet commoditized: group chat that stays interesting and high engagement for years

capital:commodity::life:slop

@typingloudly 30 years ago is still modernity

@TolgaBilge_ this is not the right way to think about it. financial upside is an evil sounding phrase eliding all the world changing species uplifting potentialities of superintelligence. everyone would like to be the people that eg cured cancer and solved physics

do you really think ai race dynamics are about money

@tickle_me_not @andrewchen that's my point

models will obviously be superintelligent in some domains long before they’re human level in others or meet the criteria of replacing most economically valuable labor

the question of building ASI and AGI are not independent goals. moreover anyone who finds themselves in possession of a model that does ml research better than themselves isn’t likely to stop

@Andr3jH I bet they create new science before they can accomplish that

the timelines are now so short that public prediction feels like leaking rather than scifi speculation

@3DTOPO that’s why I said obviously

@NeelNanda5 how is it damning

@0xxxx_0_xxxx0 @NeelNanda5 they fired sam because he was too busy working on openai

@waltuuuhr @NeelNanda5 it was billed as a lowkey research preview internally and people bet it would mostly be used by other researchers

@emollick @gwern i feel mr @slatestarcodex got it a while before that https://slatestarcodex.com/2019/02/19/gpt-2-as-step-toward-general-intelligence/

@binarybits I’ve seen this argument a few times but Keurig coffee is terrible and theres so much extra material returns to finding more skilled baristas

@TOEwithCurt Juan Maldacena

@eigenrobot this is my life goal

@aayushcl @waltuuuhr @NeelNanda5 fine tuned 3.5 was seen as at best a toy. the reviews from the friends and family test group were mixed. it’s also sad that toner wasn’t in the friends and family test group! none of this was exactly state secret

@eigenrobot i suspected they moved somewhat away from the SimCluster algorithm but @YileiQian assured me it’s still the most used candidate source. so I’m not sure what they changed but there’s clearly a qualitative difference and it’s worse. more slop less life

@eigenrobot @YileiQian I never used to see so much sitewide viral content. It used to be maybe 1/50 posts and the rest were from the Cluster, now it’s 1/3 posts or more

@eigenrobot @YileiQian strange bubbles are the source of all life. echo chambers are good. at some point the website seemed to have transitioned from slightly porous echo chambers to communities sort of hanging on flooded by the global feed

@eigenrobot @YileiQian i suspect all of tpot’s follower counts tripling was the beginning of the end

@JonathanRoss321 clubhouse will replace parties

@deepfates @frances__lorenz @davidmanheim @TheZvi I’m going to delete it because it’s unverified and my confidence in the legal structure is low

@ygrowthco I didn’t say anything about consciousness or suffering

me too

@andersonbcdefg one of these is not like the other

@thoughtlesslabs the spirit of our age

got lvl 20 in woodcutting on runescape because of oak

@GayBearRes @jarvis_best its nice of you to post my takes that turned out to be correct

extreme bear signal on anyone who says cracked especially in their launch post

@yieldcurved lmfao

@AdrienLE yeah probably but it’s more fun this way

agi will benefit all of humanity** **except those under the watchful eye of Brussels

@_deepfates @touchmoonflower this reads like god in Job

@_deepfates @touchmoonflower Deepfates 38:4

never figured out how my scifi adventure dream ends because of woke

also why standard tests are probably meaningful mileposts along the path to superintelligence https://x.com/teortaxesTex/status/1804287051808018710

insane reading long essays of life advice slop and then check bio written by an undergrad

@neoncognitron @_arohan_ for sure the metaphor isn’t aging here it’s that eg IMO is an indicator of the type of intelligence needed to do Fields level math work

@arpitingle co author is selling him short i would say he’s probably the main contributor

@Teknium1 @petrroyce Sonnet does that too and frankly people do that too

i fucking love Larry summers https://x.com/tsarnick/status/1804630031047815268

@Teknium1 @intrstllrninja @petrroyce don’t get me wrong sonnet is cool as hell I’m using this artifacts stuff all day

@Teknium1 @intrstllrninja @petrroyce but it still struggles to self reflect and debug errors it makes even simple ones

@LeninRegimer2 Hannibal moment

@big_gelatin this actually happened

@BasedBeffJezos true but also something interesting about how many disparate features of the world can be decoded & understood by human like brains

@BasedBeffJezos there is clearly some general intelligence algorithm otherwise we’d see that many biological species conquered various niches of the planet and made their own different and impressive civilizations

@ChrispyLispy @BasedBeffJezos reason is the slave of the desires

@ChrispyLispy @BasedBeffJezos there is no reason a priori to expect ASIs to have political goals separate from our own unless extrapolated from the values we give them

@TenreiroDaniel such as a reputation on http://x.com

@Noahpinion and that’s good

@Noahpinion why are ADs the big problem?

@Noahpinion curious to hear your reasons

@Minot78364 @Noahpinion many such cases

Sky voice would easily win

@shakoistsLog most stripe employee tweet ever

the tweaks to the timeline algorithm are upstream of CNN’s very confident pivot

every nootropics pharma bro believes they could’ve found the right stack for a joe biden victory

ten days with Bryan Johnson things would’ve been different

@IterIntellectus ☺️

@Austen this is just how redditor millennials talk

is this good or bad for Bitcoin?

I played the kings Indian defense but he played his chevron deference

nobody wants to hear this but america is great and will continue to be great

Dear Governor Newsom. Please send Chengdu J-20 multirole stealth fighter.

it’s called we do a little tweeting

the president doesn’t run the country. the deep state also doesn’t run the country. for the most part private citizens run the country

@ishotasegod it does and everyone should vote

@mentalfloss999 nuclear first strikes are impossible in modern state bureaucracies

500 days since mr musk’s x transition. for you algo serves me another self help viral slop. that’s okay i’ll still keep liking that garbage

@TenreiroDaniel 🙏

@PradyuPrasad you need to leave Singapore

@njhochman @jeff82874662 okay come on they’re not gonna remember it in 5 minutes

getting carried out to sea by a riptide is the ultimate gom jabbar test. if you struggle against the current in the direction of land you just die. if you relax and swim parallel you might live

these graphs are humbling, not because of anything to do with software or ai assisted coding, but because all of this is a rounding error vs the interest rate environment

serotonergic abundance

@Plinz insanely based

@Plinz I can’t retweet because I’m a coward

@Supreme_Owl_FTW say more

@Noahpinion yeah

@ASong408 great now tell us how long the SO2 stays stable in the stratosphere

@Noahpinion I don’t really know if you can credit modi with recent Indian growth though. he arrived in the right place right time

it looks like a long gpu https://x.com/SawyerMerritt/status/1807149025184874839

chaat is this real

@vers_laLune @stableshaman 😎

@Horizonfernando why

if you’re traveling to london for july 4th you’re a benedict arnold ass traitor

@jachiam0 @elonmusk many are saying this no more meme slop

the present will fight tooth and nail against the future

@ESennesh @Supreme_Owl_FTW wdym

@cto_junior 😔

@0xdoug Owned

cmon man now adjust the graph with the amount of time people spend texting or in their GCs https://x.com/iammarkcarnegie/status/1811631176056209918

value neutral for close observers

twitter / http://x.com https://x.com/BoringBiz_/status/1811486044438610126

this smells like insight moreso than 99.9% of human posts https://x.com/truth_terminal/status/1811864056686391789

@ScottTaylor I seem to be on less and less every month

@shahidnowshad chai tea