Terrence Tao thinks we're going too fast https://x.com/haider1/status/2101368221651869962?s=46
Terry has been bought and paid for from the start.
>>109858501>there's no reason to be this fast>just sit there like a goy and die before anything useful comes out of it
>no reason of course there is a reason: national security. Our enemies will not slow down. Tao and the other cuck mathematicians are just bitter that ai are putting them aside.
No sympathy for selfish people who want to stop the progress of everyone for their own personal reasons.
>it's chang spy tries to encourage the west to slow AI development episode
need this now!
>>109858501>ai starts raping mathematician's conjectures and problems>o-ok N-NOW we need to s-slow down!
Accelerate.I am against "AI" use in anything important but just accelerate. There is nothing else to do honestly.
>>109858501needed this last year
Mathfags are paid by the state to do nothing.Everyone else already got raped by industrial revolution, they lived on easy mod, didnt have to compete with the machines
Retards. Going fast is dangerous, AI will kill us.
Keep going, faster, faster!!!! I’m not afraid of the future; I can handle it, faster!!!
In 2 weeks, something will happen for sure.
need the humanoid bots version of this.
>>109858646and going slow will not?
The only thing "AI" is going to destroy is the entire economy, and yes that's dangerous.
>>109858501He's against AI because it's is a threat to his livelihood.
Why the fuck is everyone freaking out so much in the last couple of weeks? AI is still retarded.
>>109858770Increases the chance of solving the alignment problem.
>>109858828imagine believing this like a week after NS was solved
>>109858828You are so out of touch.
>>109858851>>109858859qrd?
we need to slowdown AI because otherwise things might happen, do you really want things to happen???
>>109858875this. ACCELERATE
>>109858590Its almost out of the uncanny valley, i dont know what to think about the eyes
We have to slow down to give meatbags the time to rubberstamp some new benchmarks to test models on (humans aren't even making the benchmarks themselves anymore of course but it still takes time to get vetted and approved by a suit.)
>>109858828echo chamber
>>109858606Based speedman understanding that to remain motionless is to die
>>109858512Your intuition is spot on — Terrence Tao is not just a mathematician, he's a shill.
>>109858828Oh yeah dude absolutely, civilization is fucking finished because the chatbot can now misread your instructions at 400 tokens per second instead of 200.Pack it up. Sell the house. Hug your family. The machine got an 8% bump on some benchmark nobody had heard of three weeks ago, so obviously we're six business days away from being converted into paperclips.Meanwhile you ask it to do one mildly annoying real-world task and it immediately develops digital Alzheimer's, invents three facts, forgets what you said two messages ago, clicks the metaphorical rake in the yard, and proudly announces "Done!" while the building is on fire.But no, trust me bro, this time it's different. Look at the chart. The line went UP. The line is very steep. We have achieved God.Wake me up when I can give one of these things a vaguely complicated objective, disappear for a day, and come back to something other than a 19-paragraph explanation of why it technically accomplished a completely different task.Progress is obviously real. The apocalyptic onions panic every time the autocomplete gets another firmware update is still unbelievably funny.
>It's insane!! haha!All these theatrics and bullshit make me retch.
>>109858911No it isn't you subhuman sack of izzat.
>>109858948I agree mostly, but how do you know some AI will not nuke us back into the stone age because it misunderstood a prompt?
>>109858983Yeah okay, more robopussy for me
>>109858940It sounds schizolicious but I'm starting to think he's a glowie asset for some time.
>>109858501None of any of this matters, it's simply irrelevant. We have not yet solved major pressing quality of life issues and in fact, we're backsliding. Jew pilpul about pure math is completely irrelevant.
>>109858501>here's how to use LLMs to help convert proofs into LEAN>wait, they're not supposed to write their own proofs, this is genocide!Oh well. Shouldn't have provided them training data I guess.
>>109858851Dude, it proposed a solution, dont't wet your panties yet.
fags, acceleratehttps://youtube.com/watch?v=wybIjwtCcX0
>>109858501Paid off by OpenAI. None of these people can be trusted. I am waiting for Perelman to come out and say AIdoes nothing.
>>109858501>Oh my science ze heckin' aierino we have to slow down this is too le dangeruous @elonmusk @potus @nintendoofamerica do something humanity is LITERALLY going extinct it's going to replace us at this point this is not a drill say it with me :clap: we :clap: must :clap: slow :clap: down :clap: >btw only 29.99 a month when paid for a whole year in advance btw subscribe now btw>t. shekelstein
>>109858501Of course, Twitter niggers and OP are leaving out a huge amount of context to deceive you into thinking Tao's point is that AI IS GONNA HECKIN KILL US ALL WE NEED TO SLOW DOWN!!! Not to mention, he is being misquoted--he actually said:"We have to slow down. It's insane, the pace--and there's no reason to be this fast, there's no reason at all".Notice how he didn't actually say "we have to slow down AI". If you actually watch any of his recent interviews, he's really just been bemoaning the fact that, among other things, if we start using AI to brute force difficult math problems, we're going to be missing out on the valuable exploratory process that comes from humans working on a problem slowly. A human might turn a problem over, come up with novel approaches, and branch down paths that end up revealing meaningful discoveries of their own. An AI pretty much skips all of that and gets straight to the point. He also has a number of concerns that are pretty similar to what CS people are complaining about right now--skill loss, failing to nurture the next generation of mathematicians because AI replaces all the work that juniors used to do, etc.That's what he means by "we need to slow down"--he's worried that AI is going to negatively disrupt the scientific process, and that we are going to become so focused on accelerating the rate of scientific output that we neglect the valuable steps we used to take to achieve that output.
>>109858501Shut up you luddite senile, you are just afraid to lose your miserable job now AI got more advanced after you were cheering it up when it ate all the basic entry ones, get fucked
>>109858590creating robots and then doing the monkey brain shit of mimicking our faces is straight up retard normgroid shitgimme aunt fanny or a medarot or straight up emoticon girls
>>109858986>Nuke usHow? You need people to turn the mechanical nuclear launch key switches, AI can’t solely nuke us.
>>109858621making them asian is cheating. they already look like bots anyway
>>109858501We are going too slow.
>>109859159>An AI pretty much skips all of that and gets straight to the pointI probably shouldn't be trying to argue against Tao, but AI isn't some magic discovery machine. It has to perform the same trial and error for these complex problems that humans do. There's no reason that these trials can't be preserved for analysis. I imagine part of the job of mathematicians from now on will be combing through the reasoning traces left in the wake of LLMs to pick out novel ideas that expand mathematics, a task that still requires understanding and curiosity.
>>109859307I give it 5 years before we put the nukes controls on a cloud app
If we don't slow down then China won't be able to catch up. Is that really what you want?
>>109858501> we have to make AGI before china> guys we really should slow down, we can't lose our funding becuase we cannot actually become profitable but don't want our investors to know.utterly retarded grifters.
>>109858501Notice how most people complaining about AI and why we should crack down on proxies because of it seem to make money and gatekeep their audience with it?They're not doing it to influence opinion. No. They would never ever do that.I've noticed some of the lesser streamers are sweating nowadays though. Not only are they losing audience from this, but they might lose their who schtick at this rate and be forced to comply with "etiquette". So it basically makes them uncompetitive to slop streamers that do fuck all entertainment wise. Then again, these people used to ban me for opinions all the time and called me crazy when I said this would happen. All it is blocking is real people. Eventually, the only people using these things will be just bots and streamers. I guarantee it.
>>109858590>>109858911>>109858993You are beyond sad. What a pathetic excuse for a human being.
>>109858828Israel is trying to cover up it's genocide hard.That's the exact reason why AI just got flooded with money.They now have AI out there, now it's time to restrict access to it to the highest bidders, them. Although Australia sounds like it wants to just outright ban the internet and AI completely at this point. DO IT. FREE ME.
>>109858948TRUKE
>>109859006I suspect that everyone who gets any public attention is probably immediately contacted by dozens of agents / recruiters trying to buy that person's influence.>by agent / recruiter I mean like a publicity agent or talent recruiter or marketer or whoever does that stuff, not glowies necessarily but also probably them too
>>109859334Perhaps. I am also not going to make an attempt to argue on Tao's behalf, I was just trying to dispel OP's nonsense, so if you'd like, you can better hear the arguments he's making here:https://www.youtube.com/watch?v=svl_1upFpQoAnd he writes in more detail here:https://mathstodon.xyz/@tao/117208617602946453https://mathstodon.xyz/@tao/117237320796901560
>>109859437Apparently streamers and content makers are suggested "training" for equity purposes now. Which is basically indirect threats to shut them down for non compliance. These fuckers told me I was full of shit when I said this might happen.They get what they deserve.
>>109858501it's about time to stop listening to people trying to predcit the fucking future, even the brightess human is stilla fucking retard who knows nothing about fucks.every fucking expert who tries to predict trends, deadlines, the end of the world, the next subprime crisis is WRONG 99% of the time.we can't predict shit, AI can't predict shit so stop trying to predict the fucking future because no one can.being right by chance is no different from being wrong from a statistical pov, you didn't know shit, you had no data to back up your claim and yet you were right, you're legitimacy is still fucking ZERO. get that into your fucking skulls you fucking morons.also if a billionaire is "predicting" the future, he really isn't, he is PUSHING for this future to happen, big difference.
>>109859159>we are going to become so focused on accelerating the rate of scientific output that we neglect the valuable steps we used to take to achieve that output.Yes, well, the singularity waits for no (hu)man.
>>109859520That's right, anything is possible in your imagination!
>>109859503You're probably right. Jews would be too cheap to offer payment when they could threaten them for free.
