that would be second time it happens

now you have it straight from source:

e/acc, distinct from EA, is a rival ideology about technology.

(somewhere in this mix someone called me e/acc. i am not. if you’re going to classify me in this space, i am something like a Kurzweilian h+

e/acc is a troll ideology for upsetting EA people and scamming pmarca.)

then he complains about sneerclub for some reason, damn i can’t guess why

if anyone has Rat Explainer For Normal People 101, not written by weirdos (derogatory) that would be useful i think now that what is known here for decade broke into mainstream

LLMs doing Great Things Worth Ruining The World Over in medical research

there are entire wet labs whose software stack predates openai. go and ask anyone who works in a moderately long operating lab if there’s any load bearing Thing that runs under windows xp control or older (there will be). there are some extremely limited applications of pre-llm ml in drug design, but it’s trained on entirely different type of data and it does entirely different thing and i’d say it’s not load bearing. these things are expected to be fed with new data from testing new things provided in previous iterations and if you hit something you’ll get there on your own. all of that data is noisy

unfortunately segyges is back (got 3d for posting 911/charlie kirk meme) and posted 51-post long thread on why rats are a cult, conveniently one day after zitron dropped 1h long podcast on the same topic

that’s his twitter, but he’s also on bsky. coxon might not be

the future of humanity is kink on a blockchain <- this is what rationalists actually believe

also, this bit

When she saw he was holding a shovel, she tried running away at first (you can’t imagine how creative people can get)

i have to keep in mind that this is coming from people who had problems with opening a jar

from what i understand, it’s also impossible to copy this thing and every single instance will turn out slightly different. also it will continue to learn on whatever data it is fed so you might end up with something useful, that will degrade, is on a timer and has no backups

doubling down is mandatory for acausal blackmail reasons

professionalism or any kind of plan doesn’t matter when you have thielbux (or equivalent) (guessing)

incidentally, this is also their current best shot at turning rube goldberg cash burn pit into something that’s not one

“there is a global far right conspiracy that attempts to reinvent alchemy, colonize space and build a machine god, which they intend to use to enslave everyone else. it’s informed by misreading of old scifi and racism, but it only took off after certain hp fanfic was written. currently it’s commanded by a californian polycule with a drug problem, but also counts in their ranks white house officials, and possibly some from russian government”

unfortunately this does not make a great conversation starter. whaat? a cult took over an entire government of south korea? this could never happen here

where does thiel/paypal mafia funding situation, urbit, other fashware and broader reactionary connection (yarvin etc) fit into that

also, that one oral bacteria startup

cmos logic still beats it, because main source of heat are switching losses (all these gate + parasitic capacitances being charged and discharged, and some related phenomena). relays store more energy, and that makes situation worse, even if you make them as MEMS

right? and even then you have to go slow with explanation, otherwise you just sound like a crank

i’ll also note that assembly of reply guys decided that room temperature superconductors will magically solve energy requirements of silicon-based chips (this is not how any of this works)

that’s just “machine god will figure out relativity from three video frames” in another form

arguably it’s stupider than that

and this doesn’t even go into whether pallet of lean spaghetti is in any way insightful, or does it explain anything, or does any of these things manmade proofs do

wtf is deal of that segyges person. now they’re convinced that openai will solve room temperature superconductors in silico real soon, because altman vagueposted about it

and this all based on lk-99 fraud case, nevermind that there’s no high temperature superconductivity theory, or how prohibitively computationally expensive computational chemistry is, and i guess it’s not accurate enough anyway, this all with people that do computational chemistry coming in their thread trying to explain this, then posts this

one of them also works for anthropic

also, the unfounded confidence of openai guy makes me think that they might have done that before, and some random grad might have folded much easier under this kind of pressure

and looks like there’s another case of openai thievery https://xcancel.com/ValerioCapraro/status/2097791836269977996 https://bsky.app/profile/timhenke.bsky.social/post/3mv4ouxlnsk2z

the purpose of system is what it does