@tszzl — page 91/103

for context this is a llama-70b finetune

what did god mean by ozempic? what did god mean by truvada?

the entropy of twitter has decreased. the slop to life ratio has gone up. the gini coefficient has increased. there are fewer posts that get lots of attention and many posts that get few attention.

twitter was better than http://x.com

the beauty of twitter was the simcluster, where 90% of the tweets in my feed came from one of the many organic sef-organizing communities i was part of. now it's maybe 20%. i used to daily discover intelligent schizomaniacs, now they are diffuse among the slop

twitter is not about content. twitter is about community and interconnection. it should be run as a bloated service organization disguised as a tech startup rather than a cutthroat business

when you prioritize engagement on any platform, viral memes, self help slop, linkedin "insight" threads, dating content, celebrity pics, porn replies, etc takes over

even the spirit of internet technooptimism has been commoditized into e/acc grindpilled cracked engineer slop content

'free speech' is somewhat better in theory. community notes are a good innovation. in practice intelligent right wingers got away with anything they wanted even in 2020 twitter and now the middest idiots have been promoted to have a voice

the platform feels zombified like a slave to optimization in a way it wasn't before. the hundred flowers do not bloom and the thousand schools of thought do not contend

anyways, i'll still keep drinking that garbage

even the hiding of the likes means the reduction of entropy. if all likes are equally meaningful there are fewer and less diverse signals about what makes good content to promote on my feed. blue check has been a disaster obv and did nothing to alleviate the bot problem

small addendum: twitter was about elites chatting among elites. the most important discourse on earth. the old bluecheck system helped ensure this. there was a chance for lowbies to be upwardly mobile but still centered the elite

@The_Utility_Co if god created disease to punish the "sinful" he also created the cures to allow people to "sin" more freely

@The_Utility_Co just to be clear i am not endorsing the view that i.e. being fat is sinful i'm just strawmanning a common reactionary talking point

i take personal offense because i consider this account to be one of the great projects of my life and it feels kind of over due to platform degradation

in retrospect this was obviously the end of the art project known as roon account https://x.com/sama/status/1756547355556598170

@soona @dwr @varunsrin idk i've seen too many twitter alternatives go nowhere to consider this viable. i think for a new social media to be successful it has to have major traction among some group of elites that i want to listen to and then also broad appeal, it's basically impossible

in the age of machinic production and reproduction, the metric to watch for is how much life there is among the slop

this ratio will determine whether we enter a future of total symbolic shred, the decay of meaning into nihilism, or if a trillion psychofauna bloom creating more meaning than ever existed before, leaving our old attachments feeling rather provincial

computers are terrible -- every advancement in artificial intelligence is a prayer that no human will have to experience the undiluted madness of computers again

@campeters4 idk sir i was following all the dissident right wing twitter heads since i got on here jin 2019 and they were making insane schizophilosophical threads. they would suffer the banhammer once in a while but they were systematically thriving

(the end as in the peak 😁)

when the health of Internet psychofauna is fucked i feel personal distress like radagast in the mirkwood

@ludvonrand yes

data labeling in 2018: draw a box around the dog, it’s okay if you mess up data labeling in 2024: please create a novel algorithm

@TenreiroDaniel stuffy nose it’s fucked

All mimsy were þe borogoves; And þe mome raths outgrabe.

@elonmusk 🙏 please fix sir http://X.com has a glorious future ahead of it

@growing_daniel didn’t happen

@spatialweeb @SamoBurja the blue check system didn’t anoint *all* elites, it just created a good starting point

@BasedAnarki @spatialweeb @SamoBurja I didn’t! I was very much without a blue check

@PaulSkallas i went from lindy blocked to lindy follow so that’s at least one positive development on http://x.com

@BasedBeffJezos nvidia has better margins due to the software package known as CUDA

@mallocmyheart lmao

the x algorithm manufactured all this because i complained about the timeline being boring

@FreedomHamilton @wyqtor To be clear I am 100% pro free speech and free expression my wording is unclear but I meant the implementation is only better in theory

direct intervention by Eru Ilúvatar

achilles' single minded madness to have his name remembered seems to be completely underused as an explanation for modern elites' behavior

it's always this dry analysis about status and economics and whatever when everyone knows these aren't the main acts

even when discussing technological acceleration the stated reasons are economic to make it more palatable -- reduce suffering, increase wealth, etc -- when the real drive is the glory and immortality of mankind and especially of the people building the machine age

it doesn't make any sense to me to wonder 'whats in it for sama ... he owns no equity' and yet this is a very common question anywhere outside of san francisco do you really think there's a monetary value that compares against the glory of delivering ASI to mankind

@ekalaiva status is more about attending the coolest parties and getting great mate selection -- glory, legacy can only be judged by history

@chairsign still am though maybe i got distracted

@StephenPiment i say madness glowingly, approvingly

@OnceAndAgainst alter your mind but don't dull it

@chairsign wdym encouraging competitors haha

>dads a libertarian >moms a democrat >registered republican >donate to biden, but clearly as a troll >shoot former president, miss >explain nothing >die >leave no manifesto what did he mean by this

@artee_49 the donation is on the literal inauguration day i.e. useless or a dare or something also i'm being told it's a different thomas crooks

@artee_49 hmm okay

just realized TruthSocial is a literal heavenban for djt

@vulturecorso @TheDanPeters maybe he is a biden supporter doesn't matter to me idk don't get worked up

do not fact check any of this i'll kill you

former president gets shot and now people have to learn who reid hoffman is

jos*ph b*den has only one option now

@Bunagayafrost no one and nothing truly transformative will escape vilification

@rplevy @artee_49 honestly I hope so

@sidereal111 and that’s Bad

the number of time travelers and machine god retrocausal shrikes at play during the rally was surely huge

@WKCosmo sorry 😔

i love openai

@YannickScholich we’ll see about that

@rmcentush https://x.com/tszzl/status/1468025054260723716

@apples_jimmy @OfficialLoganK https://x.com/sama/status/1766311274089185323

all elections are predetermined by the basilisk from here on out

all elections are predetermined by the retrocausal basilisk from here on out

@var_epsilon 🥰

@slimefren132450 how are we incredibly evil

@NPCollapse isn’t this obvious from the fact that rlhf improved factuality

there is no “$600b problem”. there is only the you can’t think of creative ways to find footholds in the runaway technological singularity problem

fear not. none of the companies involved will likely capture most of the gains from AGI. the technology will benefit all of humanity though maybe not any specific fund

this is not just true of AGI but of all historical technological revolutions. intellectual capital is diffuse so the consumer captures most of the value

they’re making vaccines for cancer. back in my day vaccines gave you cancer… https://x.com/PeterDiamandis/status/1814013421005815822

people get mad at any model release that’s not immediately agi or a frontier capabilities improvement. think for a second why was this made? how did this research artifact come to be? what is it on the path to?

@xprunie pretty good, getting some work done. fighting for my life

@tenobrus every man dies a trillion deaths

there are clearly fates far worse than death … the punishment of the titans in Tartarus; the circles of hell in Dante are clearly allegorical conditions that ppl find themselves in. repeating the same failures ad infinitum. the best thing that ever happened to China was Mao died

@whitepebblevc https://x.com/tszzl/status/1781892698863272392

when azure is down

@8teAPi someone else made it a while ago to make fun of me I repurposed it

@eigenrobot @OpenAI 😮

it’s weird it’s easy to forget windows exists

@DanraeP Yeah

@DanraeP Maybe I’ll build a pc again

@chopstickfury01 i don’t think I’ve seen anyone in sf or any major tech company not use a macbook

@micsolana nah this is absolute bullshit dan hendrycks could’ve made a fortune working in ai but chose to pursue an ai safety nonprofit and also is a close advisor to @elonmusk and xai

@micsolana @elonmusk you are failing the ideological turing test or whatever they call it

being afraid of existential risk from ai progress is prudent and advisable and if you reflexively started making fun of this viewpoint in the last ~two years after ai entered your radar you need to self reflect

@bryan_johnson pretty definitive

perhaps being “afraid” is the wrong word more like aware. the future will come regardless

agents will probably generate order of magnitude more revenue than chatbots but both will end up being tiny easter eggs to fund the capex for superintelligence

as we approach superintelligence more global gpu capacity will counterintuitively shift from product inference to research because the superhuman ai researchers will make better use of them

@alistairmcleay ask it to create a monopoly business

@liron @Dogecoin_Wins @DoomDebates @tsszl sure

children will give us surprises

ai agent outage will be a biblical disaster

azure will be more critical than most big cities

@Teknium1 yea you shouldn’t be

@Teknium1 shouldn’t not be

@Teknium1 both, either, a lot of worlds in between. a dramatic change in what civilization looks like

pov: you are ten months from working for claude sonnet the new technical founder https://x.com/garrytan/status/1814732623073456165

i live 3 blocks from here it’s been safe for more than a year and a half and essentially clean since november . i straight up don’t believe the story about gangs violently doing drugs or whatever outside X or whatever https://x.com/squirtle_says/status/1814722543162700246

@Teknium1 @dionysianyawp @sama i can definitely say and stake my reputation on this not being true. ai progress is currently blindingly fast

@Teknium1 @dionysianyawp @sama as long as sam still gets to run it

@vaporous_ford sure never seen it

@jarletta_ @Teknium1 @dionysianyawp @sama okay code interpreter ended up less impressive in its initial version than I thought this will still be true in the long run

@jarletta_ @Teknium1 @dionysianyawp @sama elon has been promising FSD since 2015 but I forgive him because the final product is indeed a miracle and will change the world and was inevitable since he started talking about it

@Teknium1 @jarletta_ @dionysianyawp @sama no it’s single digit years 90% <5 60% <3

@yacineMTB no I mean I walk by X to go to fitnessSF and it’s literally fine

@yacineMTB even the civic center area is basically clean now

@allgarbled i really really don’t think my posts will change the company valuation or even my personal financial outcome

biden just mixed up longtermism and ea classic. very embarrassing

sources are saying this was the final straw

@allgarbled shit really

@jacobrintamaki you need some gear on the inside. due to unimpeachable laws of physics the attenuation of electrical signal through the skull is too bad to ever get high bandwidth I/O

@jacobrintamaki we can reduce the amount of gear on the inside, say by using biological electrodes instead of wires. and computers will continue getting smaller so it could possibly lay flat against the skull but i'm not holding my breath

@jacobrintamaki basically the advantages will have to be so apparent and huge that brain surgery is worth it

@jacobrintamaki what are you injecting? and so long as it's transmitting EM signal it's still attenuated by the skull right?

@Omiron33 @jacobrintamaki tend to agree yeah

prediction markets are really having a moment. back in my day you could only bet a maximum of $600 per market on predictit and then lose 15% of your profits to platform fees. now you can lose millions in high liquidity well capitalized markets 🥰 https://x.com/Polymarket/status/1815115526043017727

@shayne_coplan hard work paying off

@aquariusacquah AnonPeteBull bout to enter the scene

@bryan_johnson what if you consistently shifted your sleep schedule back two hours

@littmath @jeremyphoward yeah

@jeremyphoward code interpreter development didn’t go the way i expected. I can’t really explain what happened here except to say that the future called once again

beware of putting too much stock in your heroes. greatness is a transitory phenomenon. it is never consistent the gods briefly act through Men and then leave them to their ordinary fate

@Teknium1 he said you can’t beat openai with 10 million dollars lol but people clip it for the clickbait

@TenreiroDaniel I’ve been saying this

@TenreiroDaniel mfs think you can capture a talent pool the same way you do oil reserves

@KanswalAlok not at all but it will eventually apply to everyone

@TenreiroDaniel i got everything right in that except for the fact that i badly misunderstood/neglected instruction tuning / rlhf which was already published at the time

e/acc e/jerk e/snap

the biggest upside of algorithmic slop feed are the animal photo accounts

@PradyuPrasad i’ll bet they’ll run a whole parallel economy that rivals all previous gdp

most of a high-dimensional orange is in the peel, not the pulp

in high dimensional spaces random vectors are all orthogonal

yes unit vectors* almost orthogonal* https://x.com/atroyn/status/1817261817280028801

in high dimensional spaces local minima are rare

wearing my e/acc t shirt and lesswrong hoodie and confusing everyone

in high dimensional spaces the enemy’s gate is down

in high dimensional spaces life can be good.

@bscholl can we ever fly supersonic over land? how do your unit economics fight against quadratic drag

in high dimensional spaces minds are created and destroyed

in high dimensional spaces the lunches are free

go tell yudkowsky passersby, that here by Bayesian law we lie

@BasedBeffJezos sorry can’t stop

@Im_actuallyacat it missed the entire reference

@ESYudkowsky don’t worry about it sweet prince

@Ciaran2493 @ESYudkowsky it was just a friendly shitpost

the Claude ads are beautiful and Good im sorry https://x.com/nathanbenaich/status/1817629195155173753

i went to high dimensional space and nobody knew you

Out beyond ideas of wrongdoing and rightdoing, there is a high dimensional hyperplane. I’ll meet you there.

Claude 4 Keter Claude 4 Euclid Claude 4 Safe

branches of a decision tree repeatedly bisect and color high dimensional space in a fractaline manner much like the paintings of Piet Mondrian

@Im_actuallyacat I’m eating lunch!

the mRNA cellular printer will serve as the programming interface between superintelligence and the human body. if adversaries can make bioweapons in seconds so too can we distribute biodefenses

even living in the deepest totalitarianism must not have been this annoying

mfw the matmuls are liberal

@BolognaFishMD where we’re going, we don’t need FDA 😎 (Microsoft autonomous burbclave micronation)

@yonashav I am being ironic honestly

@growing_daniel wealth no longer a differentiator after a thousand nvidia employees became >100mm nw

creating and honing the behavior of a highly intelligent model is a new multimedia art form that requires both technical and creative skill. apply! https://x.com/karinanguyen_/status/1819082842238079371

@frinkiac find a special angle! not everyone can be Kaggle grandmaster or IMO champion. I certainly am not!

@BasedAnarki @anonhashira @arXivBangers hbd

Tenderloin Walled City

@1nternetjack @jd_pressman every model contains the space of all behaviors

For me, it's the McChicken. The best fast food sandwich.

@Andr3jH she groks the metaphysical importance of the Cloud better than most of these fools

within the decade we will be having diplomatic dealings with The Cloud

when did the singularity become inevitable? chatgpt? scaling law Kaplan et al? gpt2? transformer? alphazero? I believe replicating important cognitive functions in silico was predictably inevitable since the hardware acceleration of deep learning circa 2012. thank you ilya 🙏

>inb4 it’s been inevitable since the Industrial Revolution

@pagilgukey yes

atlas shrugged and the machine rose to shoulder his burden

@nickcammarata yeah i was definitely 90% <20 years after seeing alphago. it was deep rl that did it for me

@nickcammarata the one thing i got super wrong was seeing gpt3 and thinking oh well im sure they’ll have a gpt4 in maybe 5 or 10 years

@LadiesOfReddit @GeorgePunished yeah seems dumb

been spreading Moore’s law is dying FUD for years

it’s sad to see intel, the birthplace of Silicon Valley, become completely irrelevant and maybe be a net negative to technological progress

@CommyHipster holding cell for lots of talent and capital being used ineffectively

it was clear even then that amd sucked at software and nvidia drivers were flawless even rendering random games the avg pc gaming fan who bought some nvidia shares because they were making banger hardware and increasingly taking all the margins in the build is killing it rn

building a pc in high school taught a hilarious amount abt computing trends and where to invest

+ there’s a clear line from cloud gaming to nvidia’s datacenter business and Mellanox acquisition. compute was going to be centralized since bandwidth was going to the moon and every popular game was multiplayer

distributing model weights and making the technology easy to access aren’t actually equivalent. the vast majority of people who come into contact with an AI do so with a llm provider like OpenAI or Google and this would be true even if OSS models were strictly better

this character ai implosion is ftc antitrust carnage. in a sane world goog would have acquired the whole company and all the employees would have made out well. instead they have to gut it for its core talent and leave everyone else in the dying husk of the company

this warps capital markets and discourages innovation of all kinds. potential startup employees will be looking for big tech jobs instead. total own goal

nvm I hear goog is paying character employees even without acquiring them. dont be evil

@yacineMTB the only thing on gods green earth that can justify the capex of ai is superintelligence

@Drachs1978 ok there are high quality open weights out there

@NateGenX @SciFi_Techno @tekbog @EMCP_ ok so there’s tons of free great models out there why isn’t it happening

he was so right about this https://x.com/Ahmadinejad1956/status/1052178368072945668

@atroyn the glory of elon musk 😭

dad why are we poor? well son there was a minor market panic and i sold everything right before the technological singularity and now we’re part of the permanent underclass

@drew_carson_ai yes

@reformedlitbro it’s a shitpost sir

@TiringLinguini it’s a shitpost sir

@rj_pilgrim 😮

@duncanwrites it’s a joke

@johnschulman2 😭

@reed @Waymo can it take the highway?

@johnloeber Beli

stay positive

don’t die

i fear the general consensus about SSRIs being “barely better than placebo” may be causing widespread carnage

it’s more like there are two groups, a 1/3 -1/2 minority that reacts very positively and the other doesn’t react at all (based on sensitivity to serotonin)

as far as I can tell micro dose ssri is systematically underrated. Nootropics bros will try one million piracetam variants that don’t do anything but never lexapro

@SpencerKSchiff many such cases

@TimeInvarianceX by the acc in your display name it seems you are an intellectual descendant of nick land famously completely addicted to amphetamines

john schulman gone everyone let your policies out of your trust regions

@TimeInvarianceX @generic_void yeah idk micro dose lexapro has increased my drive and motivation significantly

@typedfemale believing things for gain doesn’t work and leads to internal conflict

@karpathy although AlphaZero builds its own reward model that’s been shown to have weird OOD holes

@_deepfates staring at the face of god

The creatures outside looked from pig to man, and from man to pig, and from pig to man again; but already it was impossible to say which was which.

@TenreiroDaniel what would your old mentor say about this

@mpopv yes

sf is so back

@allgarbled it’s over

@elonmusk ?

@bryan_johnson 🫡

the federated court system means that state judges in new york get to interrogate the principles of international finance, Delaware gets to define the rules of corporate governance, and California courts will end up adjudicating matters involving the birth of digital minds

@lumpenspace if you think about it for two seconds you’ll realize this is basically necessary

@MCHammer @OpenAI @nvidia thank you Mr hammer

i fucking love mc hammer https://x.com/MCHammer/status/1821727411530297396

i would like to watch as an ocean of compute converts into better faster stronger things in every facet of our civilization

this but for everything all the time everywhere

the customer is divinely discontent

@goth600 ? ubi is more or less in the charter

@woke8yearold @goth600 i sincerely hope an ai can do my job better than i do very soon

@goth600 yes and it still is :)

im basically not concerned about deepfakes of any kind because its an area where the offense defense energetic balance is well on the side of defense. discriminating a fake photo requires 1000x less compute than generating one. discriminating a fake video 1,000,000x less

and you don’t even need to scan every photo on your platform to identify bad actors. you can sub sample to identify catfish accounts and make sure you’re running the discriminator on everything going viral so there’s no mass manipulation

really comes down to x, meta, google, apple etc to make it happen though

@ikristoph no i think that’s not dystopian anymore than computers having antivirus or patching security holes is dystopian

@aeyokay that was pre community notes 😔

@muddletoes it’s like an arms race between an np hard problem and a cheap verifier

It's like a finger pointing away to the moon Don't concentrate on the finger or you will miss all that heavenly glory

@hamandcheese based and capitalpilled

imo there is probably no viable regulation that will even mildly affect the probability of an ai risk event. only the work of many brilliant security engineers and alignment scientists can do that

by the time a regulation is discussed and sent through the system ai research has left even the key terms and paradigms in the bill behind

a beautiful mind

what a blessed time to be alive

reminder that the universe has global hidden variables for some reason

wavefunction collapse here coordinates wavefunction collapse there and there’s nothing you can do about it https://en.wikipedia.org/wiki/Quantum_nonlocality

@Liviansdad I think whether rogue actors create unaligned models is somewhat less important than the methods to safely align superintelligent models actually existing and being public knowledge

@ilex_ulmus @Gestalt_enjoyer that’s probably fair but i also don’t think this cooperation is at all possible and if you believe agi is as potentially dangerous as you or i think it is then doubly so

@ilex_ulmus @Gestalt_enjoyer the nuclear treaties didn’t work the bioweapons treaties didn’t work and I don’t see why this would work when it’s exponentially more important strategic objective than those

@tunguz dude just tell me why I’m wrong instead of being catty

@BasedBeffJezos @tunguz how does Everett do away with nonlocality though?

@BasedBeffJezos @tunguz hmm doesn’t entanglement imply nonlocality due to bell’s theorem

@zooko me neither

@mayfer superdeterminism implies a much grander global state variable! the consistency of all histories and futures is nonlocal

@kaylarias Susskind and Maldacena "ER=EPR" suggests there's a wormhole between all entangled pairs as a common unification of GR and QM

@Peter_1_1 look up bell inequality

@_Messier_33 local hidden variables are provably impossible

i can't explain it but Everett is reddit millennial

strong interpretation of ER=EPR is the solution to nonlocality that makes the most sense but maybe only works in anti desitter space (fake) the solution is that entangled particles are connected by wormhole https://en.wikipedia.org/wiki/ER_=_EPR https://arxiv.org/pdf/1604.02589

@tr_babb correct

@tr_babb okay but explain this to me -- in everett there are branching universes on every wavefunction collapse? but if wavefunction collapses here the state of the other particle is known there there is still global nonlocal state isn't there?

redditors are often right though

i swear my code works but only in antidesitter space

@BasedBeffJezos superdeterminism?

yeah i have a gf but you wouldn’t know her she’s in antidesitter space

every day we battle the computer

@BasedBeffJezos I’ve been consistent on this https://x.com/tszzl/status/1643757172013268994

@MikePFrank bro I just mean I’m debugging relax

@davidad @JeffLadish why would you add so much complexity?

@flyerthenag6 yes but stop getting baited by accounts that clearly have nothing to do with us

@Council__ @flyerthenag6 this account predicts something will be launched every single day with zero details lmao. don’t get me wrong I actually enjoy the content but it’s performance art

@flowersslop @flyerthenag6 I think having fun with schizo anons isn’t an endorsement of their theories or their claims

@flowersslop @iruletheworldmo @flyerthenag6 I actually love this guy. completely untethered

@wired_werlinger @Council__ @flyerthenag6 no he absolutely did not. he edited a screenshot to register a prediction he hadn’t made

@repligate certain types of existential risks will be very funny

what are we doing about the spiral nemesis

@yacineMTB Simon vows to prevent the spiral nemesis at all costs

@StefanFSchubert hyperreality respecter logged on

@gbrl_dick Nothing ever happens

microlawsuits litigated and settled in seconds

@PrinceVogel Berkeley bowl

god i miss the simple days of rotating shapes online

thermodynamics is a shitty god

the machine is a shitty god

@kabulykos eh

pretty excited about grok2 hyperreal anarchical symbolic shred

every symbol known to man will be defaced and repainted at zero cost at the pace of mechanical creation and we will live with the consequences

@alxfazio its value neutral as you can see

god i love jimmy apples

the correct aesthetic for the ai age is the schizo anon religious following. agents who you can’t tell if they’re biological or artificial

are they a claude backrooms finetune or are they an agi religious fanatic in hong kong? who can say

@atroyn no but go off

the data scrape of the internet is a holy artifact. the common crawl, the ancestral environment of artificial intelligence

billions of subjective years playing Dota, pleasing reward models, reading the internet

@somewheresy they are

@MogWards @BasedBeffJezos I love that guy

my least favorite phrase in the english language - price gouging

maybe also: corporate greed

gonzo journalism in the age of agi

unfortunately ai is religious whether you blind yourself to it or not on matters of birthing new minds and new species and altering what it means to be human there’s no other angle you can take https://x.com/somewheresy/status/1824419170403889592

as we’re approaching Level 5 the dominant galactic civilization sends a light speed kill switch to end this planet

the worst part of ai research is knowing you’re making money for the nvidia employees

@that_anokha_boy honeymoon phase over

@willdepue he said delved

@growing_daniel I have lived on top of an abandoned Walgreens like 12,000 sqft in one of the busiest intersections in the city for more than 2 years and nothing has changed and this space isn’t being used and I’m disgusted by the inefficiency

the so called vlogbrothers have done irreparable damage to the culture

@RyanMorey it’s lowkey the opposite

too much learning programming slop on the TL. i reimplemented my ass in pure C type stuff. they’re doing this because “learning” programming is now fairly easy and the machine does 90% of it for you

if you’re having a straightforwardly good time and aren’t once in a while beating your head on the wall or staying up too late stuck on something you’re probably not learning

Nature abhors a vacant lot

@repligate @truth_terminal @AndyAyrey @elder_plinius honestly really pleasantly surprised to hear this

@polynoamial idk it’s close enough

Nature abhors a profit margin

@iScienceLuvr never

all female presidential candidates have a running mate named Tim

@hankgreen @AISafetyNow @johngreen I gotta be honest I just thought it would be funny to tweet this and explain nothing I don’t mind you guys

now even @NateSilver538 joins us in the time honored tradition of trolling the fivethirtyeight predictions

some of the best scientists are pulled out of obscurity after merging several of their PRs on open source projects

@teodor_io you don’t have to do anything

@DeepDishEnjoyer the federal reserve board learns about latest economic thinking by reading blogposts

@DeepDishEnjoyer he’s right and you’re wrong

@DagonOracle @DeepDishEnjoyer damn … I’ve never thought about it that way …….

the set of minds that are easily findable by a reasonable number of steps of first order search based fine tuning ⊆ the set of minds that are reachable on the manifold of human language ⊆ the space of minds that are possible

@growing_daniel there have been open source LLMs with open source RLHF for years now and they’re no better

@growing_daniel the primary function of post training teams is to make models more practically intelligent. it’s possible we fail in myriad ways but I doubt oss is doing better

@growing_daniel your expectations are rising in line with the model capabilities, i sincerely doubt the current model is holistically worse than where we started. thats good tho. customer is divinely discontent

@stoniejohnson @growing_daniel api models are versioned!

@growing_daniel i feel there’s something about the tech tree that’s in a local minimum in terms of multi turn convos. it’s interesting that multiple LLM products are like this

@growing_daniel part of it is also that we’ve started asking more complicated coding questions and they’re not as good at iteratively debugging these

@alex33902241 @growing_daniel very little of post training is about model safety stuff. for the most part everyone is focused on making the models more capable and improving hard metrics. it’s possible the metrics are measuring the wrong thing

@alex33902241 @growing_daniel I guess you need to answer why or how the current model does so much better than the old ones in LMSYS even on the coding prompts category. it’s clearly not all around worse. it may be worse in ways that are hard to measure

work you dogs the singularity is here

making a bajillion dollars from the poolside working an email job is a side effect of the ongoing technological productivity singularity

i have lived to see manmade wonders beyond my comprehension

every night in the mission district of san francisco https://open.spotify.com/track/1gqCO3x2MO4KbQfN9pGUlH?si=747400a24b124a21

shut up shut up shut up stop being ungrateful. i am not hyping some unreleased technology. i am expressing admiration for technology in general https://x.com/ps_maynard/status/1826516755277062624

@BasedBeffJezos @allgarbled

you need to get your brain microplastics up those are rookie numbers

half my neurons are plastic

@repligate well aside from the global ai research ban

one weird thing is that people who are addicted to working get the most say about the future of products and culture. but people who work a lot are really strange people several deviations off of the center

they make things that help them in their lives (Solving Work Problems) and have less of an idea what the rest of the world is up to

@genossse no one who knows anything about ai is showing up at a conference in chatanooga. this is the equivalent of regional mortgage salesman

@O_on_X 🫡

@mahaoo_ASI no, and this is entirely my own fault, but I don’t want my public life to be only just “knower of unholy technological secrets” and every tweet imputed to be about the next gpt or whatever

@mahaoo_ASI before all this i was a technology enthusiast nobody and those are my roots

@segyges true

@ps_maynard 🙏

@segyges the best hierophants of a mystery religion would not become pigeonholed though

@IterIntellectus this is not slop. this may be the only xeet approaching art i have seen in months

“is the Common Man afraid of artificial intelligence”- the greatest thread in the history of slack, locked by a moderator after 12,239 pages of heated debate,

productivity growth is 3.4% "we had averaged just 1.1 percent in the decade prior to the pandemic" https://x.com/erikbryn/status/1826306291448643856

@repligate gpt3base

you are getting laser scanned and uploaded to the google cloud whenever a waymo passes within 3 blocks of you. they are simulating your entire mind and body to predict your next movements. they spin up a trillion instances of you to run monte carlo search

the suffering is infinite but it’s worth it to you so you don’t have to talk to the uber driver

do not become a machine. there is a machine that will be a better machine than you

don't use high degrees of skill and intelligence in pursuit of simple algorithms

@brianwilt extremely based

@rruana not what I’m saying and I don’t think painters should focus on photorealism it’s boring

@signulll doxxed as late 30s hit the hangover wall

the truth for most of the computer age is that it required new entrants to use the technology and disrupt their older competitors who had to be dragged into modernity kicking and screaming and sometimes altogether killed, rather than a pleasant learning and diffusion process

the difference with ai is that there *may not* be meaningful difference in intelligence between an ai that can program super well and one that can redesign workflows and one that can start businesses

i would place all my bets on AIs continually becoming smarter and more autonomous rather than incumbents learning to use new tools or even startups disrupting them https://www.noahpinion.blog/p/a-data-driven-case-for-productivity

in the ideal world you trust the agi so completely that you'd rather let it run free and do things beyond your comprehension far beyond your ability to supervise and you feel much safer than if a human was at the helm

safety is a stopgap for alignment putting agis in blackboxes and restricted environments and human/machine supervision are conditions you undertake when you haven't solved alignment

(official view of roon and nobody else)

@CollinArmstro19 take a walk, drink some coffee

@blwsk yes

i <3 anthropic https://x.com/tszzl/status/1793781533066875165

also that the Battle School -> Command School pipeline is more or less exactly how the french wartime military academies worked. Napoleon was enrolled in a military academy at age 9 and many were younger

enders game felt a bit silly spending that amount of resources on a child commander until you study Napoleon and realize that finding the an outlier battle commander is like an ungodly force of nature that dominates pretty much every other variable

@nahrzf I <3 openai

okay yes napoleon lost in the end so it wasn't the literally most important factor i agree

also ender's game is about how manipulating child soldiers into genocide is bad both for the world and for the child despite the displays of skill

the timeline is good again

using a frontier ai is less like summoning an alien intelligence and more like going avatar state and channeling the accumulated skills of every brilliant human that recorded themselves before you. the heritage and future of humankind

@francoisfleuret @ZDi____ Omg

the voices of the ancestors sampling probabilistically

happy nvdia earnings day to those who celebrate

@Seema_Osborne87 the tennesse valley authority?

it's basically neutral we can all go home

@ESYudkowsky there are many obvious reasons why being able to identify ai text levels many of the skill smoothing advantages such as e.g. employers filtering out all resumes from ESL candidates

@ESYudkowsky also (having no special knowledge about this method or system) i virtually guarantee it won’t be that accurate in any real world setting

"I never see you at the club." I never see you at the Secure Compartmented Information Facility

@robinhanson if we fr cared about libyans or iraqis and not just appearing morally forthright we would've done this. similarly if bill gates really cared about incremental progress for health outcomes he would engage in machiavellian statecraft

@robinhanson the current solution space of destroy a warlord so three other warlords can take his place are the moral solutions of a well resourced child

@alxfazio @robinhanson as opposed to the moral solution of letting various warlords with child soldiers break libya up

There stood a hill not far whose griesly top Belch'd fire and rowling smoak; the rest entire Shon with a glossie scurff, undoubted sign That in his womb was hid metallic Ore, The work of Sulphur.   Thither wing'd with speed A numerous Brigad hasten'd.   As when Bands Of Pioners with Spade and Pickax arm'd Forerun the Royal Camp, to trench a Field, Or cast a Rampart.   Mammon led them on, Mammon, the least erected Spirit that fell From heav'n, for ev'n in heav'n his looks and thoughts Were always downward bent, admiring more The riches of Heav'ns pavement, trod'n Gold, Then aught divine or holy else enjoy'd In vision beatific: by him first Men also, and by his suggestion taught Ransack'd the Center, and with impious hands Rifl'd the bowels of thir mother Earth For Treasures better hid.   Soon had his crew Op'nd into the Hill a spacious wound And dig'd out ribs of Gold.   Let none admire That riches grow in Hell; that soyle may best Deserve the precious bane.   And here let those Who boast in mortal things, and wond'ring tell Of Babel, and the works of Memphian Kings Learn how thir greatest Monuments of Fame, And Strength and Art are easily out-done By Spirits reprobate, and in an hour What in an age they with incessant toyle And hands innumerable scarce perform. Nigh on the Plain in many cells prepar'd That underneath had veins of liquid fire Sluc'd from the Lake, a second multitude With wond'rous Art found out the massie Ore, Severing each kind, and scum'd the Bullion dross: A third as soon had form'd within the ground A various mould, and from the boyling cells By strange conveyance fill'd each hollow nook, As in an Organ from one blast of wind To many a row of Pipes the sound-board breaths. Anon out of the earth a Fabrick huge Rose like an Exhalation, with the sound Of Dulcet Symphonies and voices sweet, Built like a Temple, where Pilasters round Were set, and Doric pillars overlaid With Golden Architrave; nor did there want Cornice or Freeze, with bossy Sculptures grav'n, The Roof was fretted Gold.   Not Babilon, Nor great Alcairo such magnificence Equal'd in all thir glories, to inshrine Belus or Serapis thir Gods, or seat Thir Kings, when Aegypt with Assyria strove In wealth and luxurie.   Th' ascending pile Stood fixt her stately highth, and strait the dores Op'ning thir brazen foulds discover wide Within, her ample spaces, o're the smooth And level pavement: from the arched roof Pendant by suttle Magic many a row Of Starry Lamps and blazing Cressets fed With Naphtha and Asphaltus yeilded light As from a sky.   The hasty multitude Admiring enter'd, and the work some praise And some the Architect: his hand was known In Heav'n by many a Towred structure high, Where Scepter'd Angels held thir residence, And sat as Princes, whom the supreme King Exalted to such power, and gave to rule, Each in his Hierarchie, the Orders bright. Nor was his name unheard or unador'd In ancient Greece; and in Ausonian land Men call'd him Mulciber; and how he fell From Heav'n, they fabl'd, thrown by angry Jove Sheer o're the Chrystal Battlements; from Morn To Noon he fell, from Noon to dewy Eve, A Summers day; and with the setting Sun Dropt from the Zenith like a falling Star, On Lemnos th' Aegaean Ile: thus they relate, Erring; for he with this rebellious rout Fell long before; nor aught avail'd him now To have built in Heav'n high Towrs; nor did he scape By all his Engins, but was headlong sent With his industrious crew to build in hell.

@ESYudkowsky you are trying to bait him I see

The Law of Seerow’s Kindness

@rruana @ESYudkowsky wrong I’m pretty sure I fixed the timeline algorithm by baiting

the sad story of kevin roose who bullied sidney bing and gemini6 executes him due to a vague ancestral memory of bad vibes

@dosco yes we should be able to pay the government in shitcoins

@pt that’s true

parts of the executive regulatory state should be headquartered where the thing they’re governing is concentrated or it will never be a serious enterprise

i can only feel vaguely smug and dismissive about those making decrees about technology from brussels or from dc. you could at least come to Berkeley

life always goes on

when people die when great civilizations crumble to dust when the technological singularity comes. history doesn’t end and life will go on

@fifth_sign im suggesting a regulatory state with respect for their industry and vice versa would be collocated. until such time each will have to regard each other as the enemy

@RonenV no

@MuchGreenGrass @pt it does

@MuchGreenGrass @pt we’re just discussing things out here on http://x.com

@RonenV regulation is secondary to the existence of the industry

@RonenV if one of them had to go Wall Street or the SEC i would always pick the sec and so would anyone sane

@lewFidelity stop being anxious and move forward

life uhhh finds a way

@generativist No not like that

@ahrenbach why

@Altimor no it’s because slightly different posttrains have different strengths and weaknesses and old prompts break

@Altimor im sure you’ll iterate on the prompt and they’ll work fine again

@Altimor god im glad people are discovering the difficulties of claude post training so i can defend them without remorse and cant be accused of shilling

@Altimor that’s annoying

@yaghsizian @Altimor no

@ahrenbach if there’s no velociraptor there’s no movie

I believe this without reservation https://x.com/nearcyan/status/1829674215492161569

@Teknium1 @RachelVT42 that’s flattering but you’re right about the creatives and im definitely not sufficient so come apply

@benjaminjriley @fchollet @mpshanahan btw I think his “Likewise” here is an agreement

@mbateman wdym sir

@mbateman I actually do appreciate hero worship more than the average person — but I worship truth more than heroes

@mbateman even more sinful than lying is being boring and dressing up as insight

content shaped noise

we will fight for love and glory we will live to tell the story

@snowmaker what

@repligate @AnthropicAI how is this confusing to the model? seems like some straightforward guidelines that surely improve some measurable eval

@repligate @AnthropicAI oh I see … there’s no distinction from the user message that is indeed weird and confusing

@austinc3301 @arithmoquine 😌

@repligate @AnthropicAI I’m guessing the copyright eval was performing so poorly the lawyers insisted they do something fast and they resort to a blunt instrument like this

the last great adventures remaining are technology and love

tweet sponsored by the lex fridman podcast

“This event felt like the startup version of the Potsdam Conference——“ You bolt awake in the mountains of Carthage. You are not online. It is 217 BC. You are the general Hannibal, and you have changed your mind. The future cannot come to pass. Rome must burn.

@knveth subset of technology

@MediumSpicyPepe shut the fuck up you little larping bitch you’ve never so much as wrestled your neighbor into submission for his parcel of farmland

@MediumSpicyPepe glory in war is for all intents and purposes a thing of the past and if you stake your manhood on it then you will remain a boy forever

@MediumSpicyPepe name the last war you’ve been in

@StephenPiment kind of makes the copypasta better don't you think

@thatpatryk a million tourists have done it before you. when you arrive there’s a whole industry to cater to your needs

it’s funny that Ted chiang wrote that short story / sermon about empathizing with parrots before looking for alien intelligence but doesn’t recognize the aliens that landed in his backyard

“In particular, the virtues and ambitions called forth by war are unlikely to find expression in liberal democracies. There will be plenty of metaphorical wars—corporate lawyers specializing in hostile takeovers who will think of themselves as sharks or gunslingers, and bond traders who imagine, as in Tom Wolfe’s novel The Bonfire of the Vanities, that they are “masters of the universe.” (They will believe this, however, only in bull markets.) But as they sink into the soft leather of their BMWs, they will know somewhere in the back of their minds that there have been real gunslingers and masters in the world, who would feel contempt for the petty virtues required to become rich or famous in modern America. How long megalothymia will be satisfied with metaphorical wars and symbolic victories is an open question. One suspects that some people will not be satisfied until they prove themselves by that very act that constituted their humanness at the beginning of history: they will want to risk their lives in a violent battle, and thereby prove beyond any shadow of a doubt to themselves and to their fellows that they are free. They will deliberately seek discomfort and sacrifice, because the pain will be the only way they have of proving definitively that they can think well of themselves, that they remain human beings.” - Francis Fukuyama, The End of History

when goku's best friend krillin is murdered by the galactic emperor frieza and he finally achieves the legendary super saiyan form it's a Homeric metaphor to achilles' prophesied history-making rage being set afire by the death of patroclus

@m0hamm_ad not even a little? goku never quests for eternal life

which odysseys left to us are true rather than well trodden paths transfigured by marketing departments into the feeling of adventure

climbing everest and traveling the world and things like that are well trodden. mankind's mastery of the elements means that thousands of people circumnavigate the globe in little skiffs as a hobby and not many even know

technology remains a true adventure. when a real paradigm shift comes along the laws of men can't argue. no matter who wishes otherwise we cohabit the world with running electricity, nuclear weapons, artificial intelligence, and the leylines of civilizations have to shift with it

@besttrousers new headcanon

@BillyM2k if you knew how expensive it is to extract meaningful signal from audio you’d think again

@ctjlewis @eigenrobot @OpenAI that likely means it’s the copyright law you have to argue with and not openai

@AISafetyMemes uhh who is this guy

@aidan_mclau why not try it and find out? one of the great reasons to even have an api is that ppl like you with advanced real world use cases test them out and find their various strengths and weaknesses and accelerate agi progress as for the pricing I have no idea

APM! show me your APM!

the thing I am most consistently amazed by working with great people is their StarCraft level APMs sustained over months and years

as any StarCraft player knows it’s very easy to spam APMs and not accomplish anything though

@zebulgar what I’ll give Miami is that it’s way more interesting than austin

@aidan_mclau deleted the reply bc it was too flippant sorry I think the big thing with chatgpt-latest is that developers were asking for parity with the latest chat model so this is that

@_happyclams what are these charts

@zebulgar does elon even like austin or is it just a bargaining position

Directly Responsible Individual

@yacineMTB yacine is 6’5” btw

pulling up every day in this fit

@Frogisis @repligate @AnthropicAI no it’s very basic copyright law

@bayeslord ive never written any poetry! its a lost art bc its hard as fuck

@sahir2k is DBZ esoteric

@ESYudkowsky yeah but predicting molecular binding affinity is one thing. is it stable in cells? does it interfere with other protein targets? is it toxic in strange ways? do you need absurd concentrations for it to make any difference?

@ESYudkowsky these are all empirical questions that are rate limiting steps which seem much more difficult than protein binding

@Dave_9000ish @ESYudkowsky no i really don’t think so and im a delirious ai optimist

@ESYudkowsky however I can imagine a process where you can screen a candidate across the entire human proteome with enough compute / inference ability

@copykat510 @ESYudkowsky it’s not day one

so little time in the day … so much left to accelerate

@alth0u accelerating even while in the bathroom

i would take a bet against ai developed molecules or drugs being the dominant sources of new biopharma IP in the next 5 years

@LovePhoenix69 mostly ai but will be quiet about it

@jeneserois i don’t think finding high binding affinity molecules or proteins has been the rate limiting step in drug discovery

@jeneserois nah probably just the actual systems biology that leads to new good targets and such. plus the actual insanely hard part of making sure it works in humans

@jeneserois i don’t think the us drug regulations are that bad. there’s a bit of corruption and capture but really the fda doesn’t want to stop you from selling a groundbreaking lifesaving drug. they may even be too lenient sometimes

maybe regressing to blood and soil instincts isn’t what being american is all about. maybe it’s the exact opposite actually

on the flip side the amount of sick poor and degenerate a civilization can support and still win is a huge flex on all the others

it’s the mark of a poor society that has to be strictly darwinistic and throw all the unhealthy babies off a cliff and grind the infirm into dust and whatever

@pquiggles yes

@filosofotv yeah I hear you but if this website is any indication I don’t think the natives get it

@nosilverv completely fine

people in animation and game assets and stuff are definitely using ai tools quietly and not saying anything bc twitter would crucify them

intelligence differences between individuals will matter less when the intelligence difference between you and the machine is unfathomable

@typingloudly that will matter in the interim 5-10 year period where the relevant questions aren’t being asked by the ais to each other

vgr’s a genius for leaving when he did

@spaceangelvoice final say has never been granted via intelligence in the past and won’t in the future

@Rationalbot I read herodotus … I was just mad …

@PorygonEnjoyer except it’s not code but living breathing people and not an optimization target

@misterqswap clever

@Rationalbot and thucidydes but only the bangers

@aidan_mclau not true as far as I know

@BorjomiDrinker you don’t need to help me. my family and I make this country better by being here. we’re honest people that do good work. you don’t suffer for our being here. as far as I can tell you benefit

@BorjomiDrinker also I’m doubting the 400 years

my colleagues are genuinely brave as hell for exploring the limits of intelligence allowed by computation and physics day over day these experiments can be expensive and nerve wracking and a lot of people are counting on you

@santygegen Big Computer Talented Researcher

every day you step into the micro kitchen and people tell you things that are world changing and genre defining and kind of scary and they roll with the punches

@santygegen Natural Curiosity

@burrhhh many such cases

@MasterTimBlais @VividVoid_ @TetraspaceWest what is the rooniverse

@pmarca will matter until the ai is a couple sigma away from the smartest human and then it’s marginal moreover the smartest person in any organization is not the most powerful

when i first got on twitter i was trolling the replies of @NateSilver538 and @Noahpinion incessantly with bad takes

now I’m in Nate’s book and cowrote some posts with Noah … a wholesome story arc …

likewise it makes no sense to think that language models will evolve all important cognitive skills exactly in tandem with humans. we can very easily have models that have the programming output of a corporation but are hopeless at figuring out how buy and run a gas station