@tszzl — page 26/103

2020-12-01 → 2020-12-17 · posts 12501–13000 of 51,350
@_AlastairX_ unless the originator country is genuinely oppressive then the new one may find expats who are all kinds of very smart religious nuts and whatnot
2 ♥ · x.com →
if even the great Jeff Dean is a Simp then what hope have we 😞
34 ♥ · x.com →
· ↳ reply to @AndrewCutler13
@AndrewCutler13 @browserdotsys @halvorz imo the monopolists live in constant fear that their whole ecological niche will get destroyed by the next big thing, so they try to aggressively preempt it. let’s not give them too much credit though. DeepMind may be the only successful alphabet research venture in a long time
4 ♥ · x.com →
very suspicious of spaced repetition. there’s gotta be a hidden cost somewhere. maybe im giving myself future Alzheimer’s
50 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
@powerbottomdad1 maybe new synapses are paid for by destroyed synapses elsewhere just a vague intuition
8 ♥ · x.com →
"skin in the game" is an intactivist dogwhistle
26 ♥ · 1 RT · x.com →
tipping my autonomous driver
30 ♥ · 1 RT · x.com →
folx, any of the individual faces generated by the AI can’t be de anonymized. but when you post 12 of them ...
27 ♥ · 1 RT · x.com →
· ↳ reply to @piotr_wasik
@piotr_wasik the thing is that I’m quite confident that you can build new muscle mass by stressing the fibers but there need not be any similar mechanism for the brain it certainly isn’t getting any bigger
Poast your SAT and onlyfans global ranking
39 ♥ · x.com →
· ↳ reply to @eigenrobot
@eigenrobot hmmm the graph can be explained by lots of publication bias and no p hacking in that case we've thrown out some quite useful results due to an arbitrary heuristic but doesn't yet turn this all into nonsense
21 ♥ · x.com →
· ↳ reply to @tszzl
@eigenrobot consequences being that meta reviews or meta papers are more or less useless
13 ♥ · 1 RT · x.com →
met up with @harveykrishna_ today. after getting coffee with him he told me “I have become Death, the destroyer of worlds.” and took a glowing many armed form. Didn't know what to make of it.
61 ♥ · 2 RT · x.com →
· ↳ reply to @_djpn
@_djpn lol, feel like it would be pretty worthwhile investment bc it could de anonymize a whole set of ppl
1 ♥ · x.com →
· ↳ reply to @eigenrobot
@eigenrobot production crew showing when Voyager 2 hits the edge of the Vault of Heaven so they can incinerate it
6 ♥ · x.com →
· ↳ reply to @yhdistyminen
@koaleszenz @ProfoundProfit @AOCummies this is a silly strawman bc the comparison is between "reasonable lockdown and paranoia induced social distancing" vs "alien invasion level martial law" theres no reason that the former would allow the rona to reach herd immunity, especially given the vacc timeline
4 ♥ · x.com →
@browserdotsys this would destroy the larp of internet debate as a productive activity so there is no demand for it but we should make it anyway for the lulz
11 ♥ · x.com →
When vaccinations begin: • Attend all social events you are invited to • At "essential" places like the grocery store, aggressively spit in peoples mouths to spread your newfound immunity • Find indoor bars where you can sit in optimal breath-sharing distance from others
40 ♥ · 1 RT · x.com →
Let’s skip the enduring fundamental questions & move onto witty repartee and position-taking.
50 ♥ · 2 RT · x.com →
· ↳ reply to @tszzl
A LIFETIME BITCH
30 ♥ · 1 RT · x.com →
· ↳ reply to @eigenrobot
@eigenrobot anima is best introduced as the one who got NIPS changed to NeurIPS
30 ♥ · x.com →
finally some good news!
77 ♥ · 3 RT · x.com →
· ↳ reply to @antoniogm
@antoniogm basically every finance junkie spreadsheet pusher on earth is hooked on adderall because these thought cycles are not "natural" at all even for smart people
30 ♥ · x.com →
@AOCummies they are so much better than our journos at poasting the war is already lost
23 ♥ · x.com →
chicomms have already won, start learning mandarin
35 ♥ · 2 RT · x.com →
· ↳ reply to @eigenrobot
@eigenrobot hmm idk we do quite a bit of this on my team only bring a change to experiment review if theres a statsig positive change, otherwise we kill it in the dead of night
12 ♥ · x.com →
· ↳ reply to @tszzl
@eigenrobot but ranking probably works somewhat different than product growth
2 ♥ · x.com →
· ↳ reply to @eigenrobot
@eigenrobot makes total sense for projects that are heavy enough investment to require PM support but when i tweak 2 knobs and the metrics change isn’t statsig I probably won’t waste my DS’ time with it
1 ♥ · x.com →
· ↳ reply to @AlexGodofsky
@AlexGodofsky @eigenrobot yes we do this often sometimes shipping something is an engineering necessity but politically difficult if it’s metrics neutral/negative
7 ♥ · x.com →
there have got to be better uses for ML engineers than optimizing recommender systems
110 ♥ · 2 RT · x.com →
@erictrimbs Rationalist cope Modernist cope Logician cope
6 ♥ · x.com →
· ↳ reply to @techsasbro
@ne0agent1c ime autists love to systematize morality and kind of freak out when they can’t
5 ♥ · x.com →
· ↳ reply to @BrickOverton
@BrickOverton the only viable outcome is the dual state solution and the settlements are pretty much lighting that possibility on fire
5 ♥ · 1 RT · x.com →
will never forgive the MF cell that that realized it could partner up with other cells
95 ♥ · 8 RT · x.com →
· ↳ reply to @tszzl
@BrickOverton with Israel’s current aggressive posture there’s absolutely no room for cooler heads to prevail in Palestine I’m not blaming them for the situation, but they are also burning the only way out
2 ♥ · x.com →
what are the most overrated classics you've read? mine have to be Heart of Darkness and Gatsby
58 ♥ · 1 RT · x.com →
@LeefVan seems to me that english classes use them to examine certain time periods and prevailing sentiments (they serve very well for that purpose) but they are not great works of literature very boring, not super intense as philosophy or fiction
8 ♥ · x.com →
@ChainlinkStake this is just bottom grade YA i'm not sure if its been psyoped into being considered a classic already
17 ♥ · x.com →
· ↳ reply to @Andr3jH
@andrej_haulis shakespeare is very very good imo bit abstruse without attached line by line commentary though
4 ♥ · x.com →
· ↳ reply to @alth0u
@alth0u unfortunately it seems to manufacture lots of high iq worker bees with huge pain tolerance but not necessarily strong people
2 ♥ · x.com →
hmmm
23 ♥ · x.com →
· ↳ reply to @AOCoomies
@AOCoomies yeah but manufacturing a dose of vaccine is different than validating it enough to give it to ppl
2 ♥ · x.com →
· ↳ reply to @R1Jack
@R1Jack @AOCoomies yeah lol i think they have to feel reasonably safe about the vaccine before applying it obviously not the ridiculous standards of the west, but some level of safety guarantees
· ↳ reply to @jerrrrrrryyyyy
@jerrrrryyy_ looking at it more cynically, leaving the whole west in a handicapped state probably has its advantages
1 ♥ · x.com →
· ↳ reply to @jerrrrrrryyyyy
@jerrrrryyy_ what's the mechanism of distribution for this? they vaccinate at home and then send the kids abroad?
so sad to hear when great artists are kicked out of their own projects and their work stolen :(
16 ♥ · x.com →
@dcbfca one of the most transparent examples of our deranged culture wars destroying real world value
1 ♥ · x.com →
· ↳ reply to @archiveOfAwe
@_vivalapanda general management stuff and yeah beating inertia, restructuring deparments, etc you pay for the instincts of the senior consultant, not the 22 year old's powerpoint
1 ♥ · x.com →
the two richest people in the world own competing space colonization projects 🤔
164 ♥ · 7 RT · x.com →
· ↳ reply to @tszzl
cyberpunk af
51 ♥ · 1 RT · x.com →
@browserdotsys if im any example i used to venerate science and was extremely surprised when my garbage paper(s) got published so yeah had to lower my expectations all around
9 ♥ · x.com →
absolutely depressed attending neurips virtually
24 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
@rickyflows completely mogged by the academic-industrial complex
5 ♥ · x.com →
· ↳ reply to @mattparlmer
@mattparlmer forget communication, a truly advanced alien race should have a significant physical presence at least across the galaxy
9 ♥ · x.com →
· ↳ reply to @conorduffy_7
@conorduffy_7 @mattparlmer even outside the realm of Dyson spheres and such, i would expect whole planets to be littered with alien artifacts self reproducing robot swarms, etc
2 ♥ · x.com →
· ↳ reply to @edavidds
@edavidds @wannabegroncho lol >barren rock if you can't convincingly state the bull case for colonizing other worlds, it's hard to take this seriously by the way, the economy is big, has a lot of rich people, and can pursue multiple goals at once
4 ♥ · x.com →
· ↳ reply to @tszzl
@discourseloverr if it were true the immune response to covid would also make women infertile
6 ♥ · x.com →
it seems openAI is 100% badge carrying rationalists who read SSC and yud
85 ♥ · 1 RT · x.com →
· ↳ reply to @alth0u
@alth0u strange bc i think yud hates openai lol
9 ♥ · x.com →
· ↳ reply to @tszzl
@alth0u currently listening to the gpt3 guy discuss gwern
5 ♥ · x.com →
· ↳ reply to @alth0u
@alth0u their opinion is that the BPE (byte pair encoding) problem gwern was pointing out is a non issue
3 ♥ · x.com →
· ↳ reply to @mattparlmer
@mattparlmer @conorduffy_7 if someone is messing with energy output of whole stars the mechanism shouldnt matter; we should be able to sense the disturbance not like comms, which can be as covert or esoteric as the aliens desire
2 ♥ · x.com →
· ↳ reply to @james_barton
@james_barton @powerbottomdad1 I think that’s mostly PR stuff They still maintain plenty of highly confidential technologies Ask them to open source their AI inference chip and see what they have to say
2 ♥ · x.com →
· ↳ reply to @alth0u
@alth0u related: people who become deaf after theyve initially learned to talk kind of degrade in quality of speech maybe speaking requires an active listening and correction loop?
2 ♥ · x.com →
· ↳ reply to @alth0u
@alth0u lol that’s nuts hmm can we speed up TTS training with this
1 ♥ · x.com →
lol my gather town poster is adjacent to an OpenAI one bout to get all sorts of accidental traffic
30 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
nvm this backfired
6 ♥ · 1 RT · x.com →
· ↳ reply to @NeoLibBen
@NeoLibBen every day that a condo remains unrented you pay a vacancy tax of the opportunity cost therefore it's only vacant if it's not worth it to rent out why add these arbitrary inefficiencies except for fearmongering about chinese real estate buyers lol? instead: build housing
10 ♥ · x.com →
· ↳ reply to @NeoLibBen
@NeoLibBen doesn't add up at all just because you're using the condo as a financial instrument doesn't mean you wouldn't put it up for rent -- i know plenty of real estate speculators who get monthly checks you don't put it up for rent when the cost of renting is higher than the profits
2 ♥ · x.com →
· ↳ reply to @NeoLibBen
@NeoLibBen if your claim is that there are a bunch of irrational real estate speculators making unwise decisions not to rent when they should then they'll probably get a nosebleed at how fast the insanely competitive housing markets destroy them
· ↳ reply to @NeoLibBen
@NeoLibBen @BrickOverton ?? if buyers are less enthused about construction, then the demand curve goes down and developers produce less quantity at a lower price
1 ♥ · x.com →
· ↳ reply to @NeoLibBen
@NeoLibBen you have not explained why foreign investors are irrationally less willing to put places up for rent than domestic investors
· ↳ reply to @axhult
@big_shrug @NeoLibBen you're simultaneously arguing that they don't care at all about the massive opportunity cost of lost rent (thousands of dollars a month) but will care about the tiny vacancy tax (<5% rent price generally) you're about to slap on -- the logic is very bad
· ↳ reply to @tszzl
@big_shrug @NeoLibBen not to mention if what you're saying is true about multimillionaire investors, they'll be owning properties en masse and losing $1000s a month on several properties at once. renting out a property takes no time - you give it to a property management company
· ↳ reply to @eigenrobot
@eigenrobot @halvorz @sgodofsk @cwage ya this is the real issue iirc Chinese hypersonic anti carrier missile range is longer than fuel capacity of F-35 therefore aircraft carrier has to park further from shore than the bombing range of F-35
8 ♥ · x.com →
· ↳ reply to @BroductManager
@updogligma these guys seem well connected, ex MIT and stuff. Honestly I had never seen one of these before
7 ♥ · x.com →
@provinghuman GPT3 pretends to host all possible egregores But really carries none
@provinghuman on some level GPT-3 really is pretending: it’s intelligent, but it’s objective is to autoregress language. It’s not concerned with “true beliefs”. It’s only opinions are on how best to reproduce human written text
still cackling at Dara renaming Uber's war room to "peace room"
77 ♥ · 4 RT · x.com →
· ↳ reply to @alth0u
@alth0u @ProfoundProfit @RSandninja point being there’s an idiosyncratic political culture and also successful ruling 20 million says nothing about how it would work for a billion Chinese would you agree?
1 ♥ · x.com →
@browserdotsys i feel you can basically tell by, when you hit the last rep (eg your final dumbbell row in the final set), could you do one more or not if yes then weight may be too low if not then that's real progress
3 ♥ · x.com →
broke: free trade with china is good as it will democratize them woke: free trade with china is bad bc it empowered a dictatorship bespoke: the US cultivated china as controlled opposition to enforce military discipline in the 2020-30s in order to stave off decadence at home
32 ♥ · 2 RT · x.com →
insane how simple some of these top tier papers will look and then you talk to the authors and they tell you it took 2 years of engineering
26 ♥ · 1 RT · x.com →
· ↳ reply to @archiveOfAwe
@_vivalapanda but I will say the tech elite make for exceptionally bad caretakers of their surroundings autistically focused on global impact instead of getting on the MET board
2 ♥ · x.com →
· ↳ reply to @L0m3z
@L0m3z last minute swings in futures markets tend to be questions of liquidity and frenzy rather than integrating all information
5 ♥ · x.com →
the emotional toll of losing/breaking something worth $1000 vs making a bad trade that loses a $1000 are not even comparable
68 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
the former would put me in a bad mood for a week and the latter is like "ah well fuck it thats options for you haha"
47 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
@mattparlmer Airbnb product is slightly different than info ... a nomadic premium mediocre life I guess? Palatable to PMCs to pay the premium for the vibez
3 ♥ · x.com →
· ↳ reply to @mattparlmer
@mattparlmer yeah, more abstractly they’re paying for the Airbnb brand guarantee, which feels better than “I’m staying at a strangers place for a week”
1 ♥ · x.com →
· ↳ reply to @ExiledInfoHaz
@RokoMijicUK the number of right wing journos graduating top unis is probably tending towards zero
3 ♥ · x.com →
Wall Street Silicon Valley 🤝 building systems that are too complex for them to understand and eventually ruining the economy
226 ♥ · 9 RT · x.com →
· ↳ reply to @tszzl
Wall Street Silicon Valley 🤝 your dad tells you “people used to build real things in my day”
50 ♥ · 2 RT · x.com →
· ↳ reply to @tszzl
Wall Street Silicon Valley 🤝 misguidedly putting too much faith in supposedly superhuman oracles such as “AI” or “asset prices”
47 ♥ · 2 RT · x.com →
· ↳ reply to @jeff82874662
@jeff82874662 i don’t have enough insight into the motherland tbh they value intelligence extremely highly but so does eg China
7 ♥ · 1 RT · x.com →
· ↳ reply to @jeff82874662
@jeff82874662 more importantly i wanna have a convo with the abstention. wtf is that guy up to. my mans didnt read the report and he's just like "hmm i have no opinion"
5 ♥ · x.com →
one thing that makes me instantly suspicious of someones whole epistemic model is when they say "oh x person could never have real success in business, discover the truth about Y scientific phenomenon, etc due to their ideology" this is almost never true
32 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
in nearly every case there are once in a generation geniuses among your ideological outgroup another cope is narrowly redefining ideology to exclude certain successful individuals
18 ♥ · 1 RT · x.com →
rats, postrats, prostates locked in eternal epistemiological battle
127 ♥ · 6 RT · x.com →
thank u,,, not a moment too late,
31 ♥ · x.com →
· ↳ reply to @ipsumkyle
@blyyyye I honestly don’t remember it and I’m frantically trying to recover my college board password
· ↳ reply to @qorprate
@qorprate same feeling w people compulsively putting stuff in anki decks
6 ♥ · x.com →
Pedro Domingos is the fucking MAN
33 ♥ · x.com →
across the ideological spectrum, everyone hates Freud. so I will become a Freud stan
36 ♥ · x.com →
· ↳ reply to @gbrl_dick
@gbrl_dick yes exactly at least he gave me some interesting shit to read
1 ♥ · x.com →
· ↳ reply to @inspireprag
@inspireprag ridiculous amounts of space, quiet & comfort, great schools, the ability to own one's own fiefdom, generally high trust neighbors, etc
14 ♥ · x.com →
· ↳ reply to @maxisawesome538
@maxisawesome538 @inspireprag depends on the neighbors was tight knit with mine growing up, and ppl were very friendly about letting you borrow things, watching your house while you’re away, etc
3 ♥ · x.com →
the lex fridman/Joe Rogan problem where they’re uniquely high profile enough to get brilliant and weird guests but too dumb to ask the right questions
114 ♥ · 1 RT · x.com →
· ↳ reply to @abelianraisin
@abelianraisin at least don’t ask it in such a trite and unsubtle way lmao you gotta really dig in if you don’t want the perfunctory sound bite
6 ♥ · x.com →
· ↳ reply to @alth0u
@alth0u @_AlastairX_ the problem comes when you want lex to really criticize their arguments and steelman the opposing perspective e.g. waymo vs tesla and he's kind of unable to
4 ♥ · x.com →
· ↳ reply to @tszzl
to b clear they are both 100000x more talented than me at interviewing and their lack of incisive questions will never hamper their commercial success and they're pretty smart. they just can't go toe to toe with the worlds foremost experts and tend to back down easily
30 ♥ · 1 RT · x.com →
“Eric Weinstein just dropped a theory of everything” - lex, completely serious
46 ♥ · 2 RT · x.com →
· ↳ reply to @DistractedAnna
@DistractedAnna it’s also a mistake of rationalist decision theory to think that human desires are fixed rather than dynamic
16 ♥ · 1 RT · x.com →
the best social science research of our time happens inside social media cos and dating apps and such
32 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
kind of based
36 ♥ · 1 RT · x.com →
· ↳ reply to @eigenrobot
@eigenrobot lee kuan yew believed it was blindingly obvious that there would not be democracy in china anytime soon due to fundamental differences in political philosophy
3 ♥ · x.com →
say something nice about him
69 ♥ · x.com →
· ↳ reply to @aharonfriedman
@aharonfriedman ok but listen for someone who frames himself as an ML/computing podcast host he is woefully under informed even about the basics
1 ♥ · x.com →
· ↳ reply to @tszzl
@aharonfriedman pros of this: the casual listener gets to listen to George hotz explain what a hypervisor or ethereum gas limit is cons of this: he is not able to really critically analyze any of his guests arguments or poke holes
2 ♥ · x.com →
· ↳ reply to @alt1na1
@alt1na1 “why do the proles deign to talk to me”
1 ♥ · x.com →
· ↳ reply to @alth0u
@alth0u like telling your recruiter you’re fielding other offers without giving details
7 ♥ · x.com →
· ↳ reply to @antoniogm
@antoniogm definitely not that simple imo makes sense in a perfectly rational market but irl the value of ipo narratives is major
5 ♥ · x.com →
1) second brain stuff is an improvement in technology and may actually augment your cognitive abilities like writing originally did so long ago 2) its not at a point where its going to represent a step change in your capabilities and you can easily compete with a moleskine
13 ♥ · 1 RT · x.com →
its insanely hard to see through the veil of Alphabet pravda propaganda factories. there will be a huge stock price reckoning when the chickens come home to roost and we see how far Tesla is ahead of Waymo
33 ♥ · 1 RT · x.com →
· ↳ reply to @jeff82874662
@jeff82874662 pump & dump all day jeffrey. if you know what's smart you'll steer your fund to buy $TSLA
7 ♥ · x.com →
· ↳ reply to @incondrulity
@Human9b6c432f this is one of their major future growth investments and I think failure to deliver after sinking this much cash on it will be very bad PR for them
· ↳ reply to @archiveOfAwe
@_vivalapanda SV people tend to be much better versed in classics and philosophy than literally any other line of work due to the intellectualizing of the field “greed is good” is not enough for anyone here
8 ♥ · x.com →
· ↳ reply to @tszzl
@_vivalapanda Zuck’s two daughters are named Augusta and Maxima lmao
8 ♥ · x.com →
· ↳ reply to @archiveOfAwe
@_vivalapanda yep and to be honest I think the SSC/SV philosophy is very stale and running on its last legs and is still miles ahead of these disaffected polisci major status grifters
3 ♥ · x.com →
· ↳ reply to @gbrl_dick
@gbrl_dick and this isn’t to say there’s any lack of vacuous pseudo philosophers *cough naval* traipsing about the SF twitter sphere and giving warmed over summaries of the stoics — but the literal only reason there’s even a market for them is the adjacent one for actual tech philosophy
3 ♥ · x.com →
· ↳ reply to @gbrl_dick
@gbrl_dick lol it may be too late because Peter singer is a repeat guest at {BigTechCo}‘s very large “effective altruism” meetup
3 ♥ · x.com →
· ↳ reply to @vsbobclear
@vsbobclear I do have to admit he’s the most interesting public intellectual of the bunch
truly disgusting behavior tbh
140 ♥ · 7 RT · x.com →
· ↳ reply to @tszzl
look me straight in the eyes and tell me this has ever been about naming conventions rather than power trip
62 ♥ · 1 RT · x.com →
· ↳ reply to @techsasbro
@ne0agent1c there was a drama between her and a UW professor over something pretty silly (NIPS vs NeurIPS). She blocked everyone that liked UW prof’s tweets and is now tweeting her blocklist in a deranged thread
29 ♥ · 2 RT · x.com →
· ↳ reply to @rickyflows
@rickyflows she has me blocked on 3 different accounts but none of them ended up on the thread ... strange
13 ♥ · x.com →
· ↳ reply to @tszzl
the escalation from cringey porn joke to excommunicating dissenters
33 ♥ · x.com →
· ↳ reply to @R1Jack
@R1Jack yeah seriously she has beclowned herself and needs to go down
3 ♥ · x.com →
· ↳ reply to @IvyAstrix
@InfosecIvy @gigafelon oh guys I just realized that’s part of the UI haha the bar appears at the bottom of the iPhone screen when scrolling
4 ♥ · x.com →
· ↳ reply to @tszzl
“everyone I saw is male” even though there are obvious she/her bios
35 ♥ · 1 RT · x.com →
· ↳ reply to @ESennesh
@EliSennesh the clear implication being “here’s a list of people that need to be re-educated”
2 ♥ · x.com →
· ↳ reply to @jerrrrrrryyyyy
@jerry_dot_ai yeah and that’s exactly it — her intention is not to create a blocklist as this would be the worst possible way to do it. the point is to raise visibility of ppl to “re-educate”
4 ♥ · x.com →
· ↳ reply to @ESennesh
@EliSennesh it’s a metaphor for the kind of middlebrow premium mediocre authoritarianism we see today
2 ♥ · x.com →
· ↳ reply to @jachaseyoung
@jachaseyoung yeah it definitely seems overoptimized and we're well into diminishing returns territory
1 ♥ · x.com →
· ↳ reply to @hardmaru
@hardmaru @wileyinresearch would leaving some groundbreaking work in facial recognition unpublished prevent the use of these technologies? anyone with a gpu can reproduce these results. no -- instead, shine light on the work so folks can get started working on defenses
3 ♥ · x.com →
· ↳ reply to @eigenrobot
@eigenrobot "when the mountains are high, the emperor is far away. I am the emperor here"
6 ♥ · x.com →
· ↳ reply to @alth0u
@alth0u universally true in the east and why Indian serials only feature internecine family squabbles
2 ♥ · x.com →
everyone, please introduce yourself by telling us your pronouns, your MK ultra trigger, and your position on Israel
155 ♥ · 5 RT · x.com →
@vontallboy @dcbfca @aquariusacquah point being Pedro should be picking his fights. Maybe tweeting this while ppl are weighing which of them to side with is not smart
1 ♥ · x.com →
mfs read Seeing Like a State and Consider Phlebas and make that their whole ideology smh
33 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
@powerbottomdad1 also idk if he’s saying shun nice things so much as “look for the cheaper alternative and don’t get scammed”
3 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
@powerbottomdad1 ppl will often buy expensive branded shit that’s no higher quality than fast fashion etc
3 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
this is also basically a Conservative virtue train yourself for the production side of the economy don’t make yourself a slave to the demand side
15 ♥ · 1 RT · x.com →
@guywhits will think about it more and get back to you but basically the culture war is highly aligned across the production / demand split
2 ♥ · x.com →
making his position clear that the aesthetics of calling a woman crazy are a far bigger concern than actual sociopathy
27 ♥ · 2 RT · x.com →
mask off doomer blackpill joker sicko mode
45 ♥ · 1 RT · x.com →
· ↳ reply to @postpostpostr
@postpostpostr @alth0u government bought crops at a fixed price from the landowners who actually reap the benefits of harvest and not the serfs who labor on their fields lol
4 ♥ · x.com →
we should be thanking anima for compiling such a comprehensive list of ML people to follow!
30 ♥ · 2 RT · x.com →
· ↳ reply to @NLRG_
@NLRG_ the framing of loss aversion as a cognitive failure to be rid of is wrong and therefore lots of behavioral economics starts from a bad premise
24 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
@NLRG_ behavioral economics: humans are less logical than homo economicus and we are here to study the ways non ergodic decision theory: humans are more logical than homo economicus and we are here to study the ways
15 ♥ · x.com →
· ↳ reply to @marx_knopfler
@marx_knopfler No not whatever he’s tweeting about in blackpill mode rn on the actual matter at hand re: the conference and AI ethics
1 ♥ · x.com →
· ↳ reply to @tszzl
@marx_knopfler needless to say they have both beclowned themselves a bit but I gotta side with mr domingos on this one
1 ♥ · x.com →
· ↳ reply to @tszzl
@NLRG_ coming from an reinforcement learning background with its trajectory optimizations his criticism makes a lot of sense to me
3 ♥ · 1 RT · x.com →
· ↳ reply to @marx_knopfler
@marx_knopfler it is truly a disaster that some of the finest minds of our time are busy debating NIPS vs NeurIPS but hey that’s The Discourse for you
4 ♥ · 1 RT · x.com →
· ↳ reply to @eigenrobot
@eigenrobot lol the clickbaity article doesn’t do ole peters justice he has a simple point and doesn’t attempt to prove that “all economics is wrong”
2 ♥ · x.com →
@vontallboy imagine unironically saying “there’s no place for such opinions in ML”
6 ♥ · x.com →
· ↳ reply to @alth0u
@alth0u best argument I’ve seen that the name isn’t inclusive enough
7 ♥ · x.com →
@vontallboy love 2 play word games to destroy illustrative examples
9 ♥ · x.com →
· ↳ reply to @alth0u
@alth0u lol IIT ppl are considered absurdly arrogant even in India
9 ♥ · x.com →
does anyone not have myopia these days genetic freaks, all of you
34 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
@jeff82874662 that's my cynical take. less cynically: it lets major fossil fuel companies stay in business by attaching CCS which may slow down progress even less cynically: its a bad investment for technical reasons, and won't work at scale
6 ♥ · x.com →
· ↳ reply to @tszzl
@jeff82874662 i also think there may be a degree of capture of progressive groups by solar and wind interests, which is no longer an insignificant industry and relies heavily on subsidy
7 ♥ · x.com →
· ↳ reply to @jeff82874662
@jeff82874662 you might be able to argue that it will prolong the death of fossil fuels but the usual argument is that it feels shitty to send government R&D money to Exxon or BP or whoever who got us into this mess
2 ♥ · x.com →
· ↳ reply to @jeff82874662
@jeff82874662 agreed i'm just relaying the sentiment i don't think it's at all fair to blame the production side alone for these issues
2 ♥ · x.com →
bodied
90 ♥ · 2 RT · x.com →
· ↳ reply to @tszzl
@roydanroy also significant use in social network recommendations
· ↳ reply to @marx_knopfler
@marx_knopfler @ornintovitu @roydanroy ya lol the space of NNs is barely tapped even now plenty of low hanging fruit in particular one downside of the graph attention approach is that you lose edge information, which can be made up for by the increased parallelism
2 ♥ · x.com →
common criticism of NLP transformers is that they’re blank slatist and don’t encode chompskian universal grammars but there are in fact papers that do this and don’t perform nearly as well as the naive transformer https://arxiv.org/abs/1503.00075
22 ♥ · 2 RT · x.com →
· ↳ reply to @3ibar
@3ibar hmm I mean sure but no model is truly “blank slate” — they all have some kind of inductive bias, as that’s the whole point. The positional tokens enforce linear order, but do not spoon feed any grammar to the model
· ↳ reply to @MiloJKing
@MiloJKing Nate silver would be shamed off the site if our brier scores were transparent :(
2 ♥ · x.com →
· ↳ reply to @MiloJKing
@MiloJKing Silver would have a lot more gravitas if he lived in Vegas. Or Britain
1 ♥ · x.com →
· ↳ reply to @MiloJKing
@MiloJKing can you actually gamble in the Vegas betting markets from other states?
1 ♥ · x.com →
Ur voice is melodies dear
21 ♥ · 1 RT · x.com →
dr. anima anandkumar was handed a literally unloseable mate-in-two position that she could've easily leveraged to gain more power at NeurIPS and elsewhere but by the grace of god she still fucked it up
50 ♥ · 2 RT · x.com →
· ↳ reply to @tszzl
and now the man pedro is just shitposting carefree about gamergate and blm
32 ♥ · 2 RT · x.com →
· ↳ reply to @tszzl
@browserdotsys tbh i have no idea what happens next w NeurIPS but since it seems there's no clear public consensus one way or the other on who's in the wrong i am assuming they both get a slap on the wrist and nothing more
5 ♥ · x.com →
@_AlastairX_ @browserdotsys yeah probably true. do we know if neurips has taken action like this before, and if so what did they do?
3 ♥ · x.com →
· ↳ reply to @tszzl
@_AlastairX_ @browserdotsys reading the reddit r/ml megathread vs reading twitter is like night and day. basically everyone there is on pedro's side
7 ♥ · x.com →
· ↳ reply to @tszzl
sincerely depressing tbh love 2 watch some of the finest minds of our time tear each other apart in some of the most low stakes internecine squabbles i have ever seen
68 ♥ · 4 RT · x.com →
i don’t think the ethics section makes sense when the paper will go on arxiv and the code on GitHub regardless of any NeurIPS reviewer’s decision
26 ♥ · 2 RT · x.com →
· ↳ reply to @tszzl
in the case of the facial recognition paper with suspicious Uyghur targeting — are we really working under the notion that an American publisher rejecting this out of ethical concerns will slow down the use in China?
16 ♥ · 3 RT · x.com →
· ↳ reply to @tszzl
this is the classic debate over “deplatforming” — I can easily argue that publishing dangerous new discoveries will gain the attention of security researchers who will try to build countermeasures and continue the arms race
9 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
Since this is an issue where we may never get an empirical answer re: effects of tampering with the publication ecology, it’s best to follow the scientific virtue: publish if it’s interesting.
8 ♥ · 1 RT · x.com →
· ↳ reply to @gbrl_dick
@gbrl_dick public addresses became total PR hedging bullshit as soon as they realized that whatever they say is going to be leaked to the media instantly
9 ♥ · x.com →
· ↳ reply to @tszzl
@gbrl_dick it used to be possible to know what top brass was thinking during these Q&As bc they were fairly transparent about their various beefs. but they've gotten extremely adept at talking without saying anything
4 ♥ · x.com →
· ↳ reply to @gbrl_dick
@gbrl_dick bezos doesn't do a Q&A bc he doesnt give the slightest shit about seeming accountable haha the weekly Q&A is a very silicon valley thing
4 ♥ · x.com →
· ↳ reply to @gbrl_dick
@gbrl_dick zuck and pichai gotta cater to us pampered hippy san francisco brats! it seems like it used to be an excellent tool to keep everyone focused on the vision and keep loyalty high but now the functionality has probably morphed to propaganda session / pseudo press briefing
3 ♥ · x.com →
· ↳ reply to @marx_knopfler
@marx_knopfler yea I think that’s the unhappy compromise we’ve ended up at: we write an ethics section but it’s not actually counted towards anything so nobody spends more than 5 minutes on it. Not sure if it’s helping anything. Anima et al wanted it to be part of review process
2 ♥ · x.com →
catch me active learning at the decision boundary of my based/cringe classifier
46 ♥ · 1 RT · x.com →
GRRM set out to prove that lord of the rings, while excellent, had an unrealistically happy ending and that he could advance the genre by being more historically realistic & politically gritty and quickly realized that history never ends and therefore neither can his books
52 ♥ · 3 RT · x.com →
· ↳ reply to @tszzl
tolkien absolutely cackling from beyond the grave rn
25 ♥ · 2 RT · x.com →
· ↳ reply to @eigenrobot
@eigenrobot seems like they hastily rolled it out in a panic during election season
16 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
@eigenrobot there may also be more to the decision making than the PR guy is letting on
10 ♥ · 1 RT · x.com →
· ↳ reply to @nosilverv
@nosilverv @erin_nerung @CundillCapital my dude didn’t even ask an unreasonable question euro guy was clearly just shocked he was being asked meaningful things over email and responded condescendingly “lesson one of dealing in Europe ...”
8 ♥ · x.com →
· ↳ reply to @IvyAstrix
@InfosecIvy absolutely true, but how come they only start caring about the tax differential once the offices are closed? seems 2 me many people don’t value it that highly
6 ♥ · x.com →
the american neolib relationship to work ethic remains confusing
16 ♥ · 1 RT · x.com →
· ↳ reply to @tszzl
they'll never say it out loud but the encyclicals ("Lean In") clearly imply that people who work harder are better
15 ♥ · 1 RT · x.com →
· ↳ reply to @arnoblalam
@arnoblalam sure, but having leisure time is one of the highest crimes in neolib culture think of new york -- you gotta at least maintain the appearance of being busy
5 ♥ · 1 RT · x.com →
· ↳ reply to @_Jason_Dean_
@_Jason_Dean_ yea they will continue growing at a fraction of what they could achieve, but nothing new there
3 ♥ · x.com →
· ↳ reply to @Hellachans
@Hellachans what exactly happened w Miami and LA? Why did people move and why did people stick around
2 ♥ · x.com →
· ↳ reply to @tszzl
FTR im just as frustrated with sf as everyone else ... I just think a lot of this is marketing for the venture capital complex pumping out remote work companies and shitposting on twitter with the Miami mayor ...
16 ♥ · 2 RT · x.com →
· ↳ reply to @nkh1l
@nkh1l the snappy unbeatable Sorkin dialogue too
3 ♥ · x.com →
infosys / lambda school parallels are troubling
23 ♥ · 1 RT · x.com →
@TheDaveMadden Indian talent sourcing & H1B placement company takes extractive rents from these workers who get placed at like Cisco or wherever
8 ♥ · x.com →
@ProfoundProfit it is at this scale — the larger they get the more rent they can extract
3 ♥ · x.com →
there’s actually nothing more satisfying than watching some tactless half autistic new money idiot own entire industries full of generational aristocrats
249 ♥ · 13 RT · x.com →
Peter Thiel’s public philosophy: humans are too susceptible to groupthink, isolate urself, avoid competition, Peter thiel in practice: most of his wealth can be attributed to investing early in a global scale hivemind https://x.com/gbrl_dick/status/1339536707138404356
38 ♥ · 3 RT · x.com →