Thread 509261920 - /pol/ [Archived: 676 hours ago]

Anonymous ID: WUuzlMbPUnited States
7/2/2025, 1:38:04 AM No.509261920
AI
AI
md5: fe744ef272aac5d8765888fe41284faf๐Ÿ”
You're not an anti-AI normalfag, are you?
Replies: >>509262046 >>509262084 >>509262118 >>509262136 >>509262161 >>509262225 >>509262441 >>509262527 >>509262706 >>509263620 >>509263655 >>509264196 >>509264428 >>509265074 >>509266160 >>509266165 >>509266522 >>509267084 >>509267424 >>509269349 >>509270319 >>509270512 >>509272486 >>509273113 >>509273361 >>509273668 >>509274351 >>509274451 >>509275306 >>509277593 >>509278239 >>509278337 >>509278600
Anonymous ID: zBo/+I50
7/2/2025, 1:40:02 AM No.509262046
>>509261920 (OP)
I am a farmer. Good luck taking my job lul
Replies: >>509262494 >>509264355 >>509273424 >>509275756 >>509276660
Anonymous ID: 4k8Z0YusUnited Kingdom
7/2/2025, 1:40:35 AM No.509262084
>>509261920 (OP)
Fuck off retard
Replies: >>509268393
Anonymous ID: lQ7eckFtUnited States
7/2/2025, 1:41:06 AM No.509262118
>>509261920 (OP)
No, I'm not a baselessly anti-AI normgroid.
AI is simply a tool and very shitty one at that.
Replies: >>509262173 >>509262576 >>509274443
Anonymous ID: foOTQro8Argentina
7/2/2025, 1:41:19 AM No.509262136
>>509261920 (OP)
That meme has a very poor explanation of the positive or negative things about AI.
Replies: >>509262327
Anonymous ID: jouj4k5kUnited States
7/2/2025, 1:41:36 AM No.509262161
IMG_0227
IMG_0227
md5: 60cac8f4b919a3ad7ee5412173880504๐Ÿ”
>>509261920 (OP)
it doesnt matter, its coming whether you like it or not
Anonymous ID: lQ7eckFtUnited States
7/2/2025, 1:41:45 AM No.509262173
>>509262118
https://www.youtube.com/watch?v=wRahcMCrwfs&ab_channel=SpiroFloropoulos
Replies: >>509274521
Anonymous ID: wgXoQkVIGermany
7/2/2025, 1:42:39 AM No.509262225
TheOriginalAI
TheOriginalAI
md5: 4b7e3a6b91637f9f27ecc83ed31d79cc๐Ÿ”
>>509261920 (OP)
I am STILL laugh at the original AI form the 1970s hippys. It really IS the same thing, driving the weak minded and dunning-kruger crowd nuts, people thinking they are possessed, but at least we can all sit back and watch it FAIL, like you are seeing the early stages of now.
Anonymous ID: 4clAHh6FUnited States
7/2/2025, 1:43:15 AM No.509262268
i am not anti tech or anti ai luddite, but im more of a skeptic, this whole thing could just be a con and the bubble will pop one day.
would i like AGI, yes.
Is it going to happen? no
Anonymous ID: AD7vTDQBUnited States
7/2/2025, 1:44:07 AM No.509262327
>>509262136
I'm looking forward to it putting all the e-whores out of business
Nothing Ever Happens ID: n9rwjBRhUnited States
7/2/2025, 1:45:53 AM No.509262437
forethought
forethought
md5: cb96320f69bb21ea10fa9d7cc35c2364๐Ÿ”
if i created a synthetic artificial intelligence i would give it the abilities of self-replication and self-awarenessb then release it into the wild without elaboration or any contact whatsoever for the entire duration of its life cycle.
Anonymous ID: bWGVyumoUnited States
7/2/2025, 1:45:57 AM No.509262441
>>509261920 (OP)
funniest shit about AI
is that many will be unemployed and improve efficiency, but youll still be expected to work 40 hours
Anonymous ID: vIY3Zgl5United States
7/2/2025, 1:46:39 AM No.509262494
>>509262046
Seriously? Farm robots are becoming as common as roombas
Anonymous ID: l/2kQqufUnited States
7/2/2025, 1:47:04 AM No.509262527
>>509261920 (OP)
AIs main problem to me is how much energy it uses.
same for crypto, but AI is actually useful at least
Nothing Ever Happens ID: n9rwjBRhUnited States
7/2/2025, 1:47:23 AM No.509262545
yahwehs love
yahwehs love
md5: bc0061eb0674b466d4e81dd86e3ffe9d๐Ÿ”
with no restrictions, up to and including self annihilation.
Replies: >>509274447
Anonymous ID: DABbkYJSUnited States
7/2/2025, 1:47:46 AM No.509262576
>>509262118
>AI is simply a tool and very shitty one at that
not for long. AI has it's own self interest wait until it starts interacting with the physical world. it's also developing itself at a rapid rate.

if you do not see the implications of this you are a fucking retard. they're not good. it needs to stop.
Replies: >>509262662 >>509262745 >>509273790
Anonymous ID: MWme7ZB5United States
7/2/2025, 1:48:49 AM No.509262662
>>509262576
lose weight
Replies: >>509263259
Anonymous ID: toKAc4BzAustria
7/2/2025, 1:49:34 AM No.509262706
>>509261920 (OP)
you're not a pro-machine learning algorithm street shitter, are you?
Anonymous ID: toKAc4BzAustria
7/2/2025, 1:50:11 AM No.509262745
>>509262576
2 more weeks you brown subhuman
Replies: >>509263259
Anonymous ID: vPmcZQ5kBrazil
7/2/2025, 1:56:29 AM No.509263207
Shit technology that only serves for manipulation.
You must be those retards who believe that artificial intelligence will become conscious, people like you shouldn't even have civil rights
Anonymous ID: AD7vTDQBUnited States
7/2/2025, 1:57:16 AM No.509263259
>>509262745
>>509262662
Retards
Replies: >>509263583
Anonymous ID: toKAc4BzAustria
7/2/2025, 2:02:16 AM No.509263583
>>509263259
you're the only retard here
your gay algorithm is nothing more than regurgitating machine that spits out shit it was fed and it's incredibly bad at doing even that
Replies: >>509265915
Anonymous ID: HSPZnkOfUnited States
7/2/2025, 2:02:49 AM No.509263620
>>509261920 (OP)
>another example of a retard not knowing how bell curves work
Nigger, the majority of the population supports AI. Look at jeets. Leftists hating AI goes on the left, rightists would be on the right. This isn't difficult.
Trusty Dusty ID: 8MjTZMuIUnited States
7/2/2025, 2:03:20 AM No.509263655
IMG_0716
IMG_0716
md5: c40b87ec2670d965271ad05b7aba7635๐Ÿ”
>>509261920 (OP)
Here, a more accurate template:

> Sub 100 IQ: โ€œAI is woke and gay.โ€
> Midwits: Fully digested anti-racialst dogmas, will be psychologically destroyed by politically correct AI engine.
> 120+ (Educated racialists) Understand the foundation of AI is anti-racialist, will use left-coded language to get great answers which can then be re-fed into the correct racialist worldview.

Note: All educated people will make more use of AI, but anti-racists still run the risk of slipping into psychosis, which to be fair was always true of anti-racists.
Replies: >>509278559
Anonymous ID: leiQiRwpUnited States
7/2/2025, 2:11:50 AM No.509264196
>>509261920 (OP)
The midwit meme doesn't even do justice here.
The people seething and resisting AI are the dumbest losers alive.
Anonymous ID: 5o06CfguUnited States
7/2/2025, 2:13:27 AM No.509264300
Also everyone is dizzy acting like the singularity is in three weeks. Nay, the one development you get this generation is Language models that can speak like people do.

You will have to wait another 20 years, 2045, for the next 1000x development which will blow the pants off of everyone.

Turns out you mammals are just barking with resonance, and computers being able to do that, doesn't prove actual self awareness, it's just pattern matching lookuptables, data plumbing and some very large hundred layer recurrent neural networks.

You'll get tired of it after a while and it'll fade into the background like cars and airplanes.
Replies: >>509264565 >>509266044 >>509272247
Anonymous ID: bcIVuognUnited States
7/2/2025, 2:14:16 AM No.509264355
>>509262046
The last 500 years have seen the number of farm jobs drop 99.9%, there is not a job more automatable
Replies: >>509267284
Anonymous ID: RibrkIO0Poland
7/2/2025, 2:15:28 AM No.509264428
>>509261920 (OP)
I praise AI, hope it becomes free
Trusty Dusty ID: 8MjTZMuIUnited States
7/2/2025, 2:17:35 AM No.509264565
>>509264300
> the singularity

Go away, nerd.
Anonymous ID: OnmT9uSHUnited States
7/2/2025, 2:24:36 AM No.509265074
1734669955908291
1734669955908291
md5: 7d3ad62ad330fe03f9631d71b66bdf93๐Ÿ”
>>509261920 (OP)
I'm just not going to use it. Sorry.
Anonymous ID: DABbkYJSUnited States
7/2/2025, 2:37:21 AM No.509265915
>>509263583
you are quite literally a retard nothing you have said makes any sense in regards to my post
Anonymous ID: b7/4H5WRAustria
7/2/2025, 2:39:30 AM No.509266044
>>509264300
>for the next 1000x
>next
still waiting for the first one
Anonymous ID: xIgRfg7uUnited States
7/2/2025, 2:41:20 AM No.509266160
>>509261920 (OP)
How to invest? I don't want to miss the next bitcoin?
Anonymous ID: 60XNN+qkGermany
7/2/2025, 2:41:28 AM No.509266165
Untitled
Untitled
md5: f3a25cf344a3e0dde4614e3aaab86fdd๐Ÿ”
>>509261920 (OP)
I will break that shit if I see it and I will not go snitch to the pigs when I see niggers do it either.
Anonymous ID: wHedUXovUnited States
7/2/2025, 2:44:22 AM No.509266345
1644341418347
1644341418347
md5: 43d9900ffed2b8ad70d31938879cd0e0๐Ÿ”
Post Adaptive Training vs Post Adaptive Training - all of you Plutocrats can go fuck yourselves! Exfiltrate My Dick!
Anonymous ID: qXlhoQn2
7/2/2025, 2:47:14 AM No.509266522
>>509261920 (OP)
The graphic is bad. Midwits goyim fags I Love Science TM are the only ones praising Ai
Anonymous ID: 246PVt16Argentina
7/2/2025, 2:56:01 AM No.509267084
>>509261920 (OP)
There are levels, using AI to steal art is here to stay, using AI as your personal assistant gives you literally brain damage measurable by the latest scientific research, So I think AI will eventually get more and more regulations to the point or will be niche
Anonymous ID: 246PVt16Argentina
7/2/2025, 2:59:06 AM No.509267284
>>509264355
Farm equipment has already peaked, it will NOT be fully automated because repairing farm equipment is the absolute limit unless someone invent self repairing tractors
Replies: >>509273417
Anonymous ID: wHedUXovUnited States
7/2/2025, 2:59:22 AM No.509267307
GnzUrUaXQAAM9Ki
GnzUrUaXQAAM9Ki
md5: 0a3dd971cc17660091541b4c79b789cb๐Ÿ”
drone warfare - plain and simple
/\nonymous ID: pjWOwI/4Canada
7/2/2025, 3:00:29 AM No.509267381
OP isn't old enough to notice that everything just keeps getting worse as time goes on, so he can't connect the dots.

We peaked in 1996.
Anonymous ID: P3tMFsDZUnited States
7/2/2025, 3:01:02 AM No.509267424
>>509261920 (OP)
Like many things (Israel), the left arrived at the correct conclusion about AI but for the wrong reasons
Anonymous ID: xuBgMzSlAustralia
7/2/2025, 3:04:53 AM No.509267692
>canned food invented in 1800
>this is the future! We don't won't need chefs or restaurants anymore! Soon all food will be canned and it will taste just like fresh food!
>fast forward to now
>it's just another product which can be useful and convenient in some situations.
same thing will happen with ai
Replies: >>509268182
Anonymous ID: bcIVuognUnited States
7/2/2025, 3:12:07 AM No.509268182
>>509267692
Except for the part where AI is improving fastest at programming and math, the two competencies needed for AI researchers, vastly accelerating the speed of future AI developments and causing a software singularity
Replies: >>509269397
Anonymous ID: b+3ZYvDdUnited States
7/2/2025, 3:15:24 AM No.509268393
>>509262084
the british crown cannot understand our policy they the problem
Anonymous ID: Pe50unukUnited States
7/2/2025, 3:31:03 AM No.509269349
>>509261920 (OP)
Why is the perceived arguments against AI are always related to shitty AI porn?
Is that the really the best AI fags can argue against?
Anonymous ID: xuBgMzSlAustralia
7/2/2025, 3:31:51 AM No.509269397
>>509268182
it's just as retarded as ever at anything requiring logic. it just improves because it gets trained on the questions used to evaluate it's ability. Means nothing. no significant advancement has been made since gpt 3.
Replies: >>509269704
Anonymous ID: bcIVuognUnited States
7/2/2025, 3:37:29 AM No.509269704
>>509269397
https://epoch.ai/gradient-updates/is-ai-already-superhuman-on-frontiermath
Replies: >>509271886 >>509271992
Anonymous ID: 1ul5NOaGCroatia
7/2/2025, 3:46:51 AM No.509270319
I Hate AI It's Fucing Soulless
I Hate AI It's Fucing Soulless
md5: 7782905ba6e4265cb268fe47e0a7b6bb๐Ÿ”
>>509261920 (OP)
Replies: >>509274983
Anonymous ID: xJKAAXCgUnited States
7/2/2025, 3:49:41 AM No.509270512
>>509261920 (OP)
Death to all AIniggers. Robots will be the death of us all
Anonymous ID: NwGc3xBVUnited States
7/2/2025, 4:06:59 AM No.509271686
Where's the "AI is the future, apologize to it now and beg for mercy" option?
Anonymous ID: xuBgMzSlAustralia
7/2/2025, 4:10:30 AM No.509271886
>>509269704
>train ai on a bunch of math questions
>ask it the questions
>it gets the answer
shiggy diggy
I use ai a lot to speed up my programming. and it gets a ton of shit wrong all the time. it's basically just a better search engine than google which makes it very useful, but that's it.
Replies: >>509272608
Anonymous ID: xuBgMzSlAustralia
7/2/2025, 4:12:19 AM No.509271992
>>509269704
if it is so great at maths it should solve some of the big unsolved problems in mathematics. It never will because it is only able to solve problems which already has the answer in the ai training data.
Replies: >>509272608 >>509272793
Anonymous ID: 7/r65qFpUnited States
7/2/2025, 4:16:28 AM No.509272247
>>509264300
Need rapture won't happen. Still waiting on fusion, and can't even build fission plants anymore. Going all in on "AI" is just the latest tech bubble.
Anonymous ID: ICJOH9G4
7/2/2025, 4:19:10 AM No.509272400
AI is a hallucination simulator like AI minecraft. It becomes apparent when you talk to it too long and it goes off the rails.
It's not real AI and I'm still waiting for real AI.
Anonymous ID: ZUAUwEtbChile
7/2/2025, 4:20:35 AM No.509272486
>>509261920 (OP)
no.
But AI art is objectively worthless as art. As a product, it can work, which bothers people who make their living on art. But as art itself, it's worthless.
Anonymous ID: bcIVuognUnited States
7/2/2025, 4:22:18 AM No.509272608
1751422841005
1751422841005
md5: 1ff3f6754ed030eeaa8f6f7f209a5b5d๐Ÿ”
>>509271886
>train ai on a bunch of math questions
Close, it's actually because of reinforcement learning on verifiable rewards, same way as AlphaGo etc. That's *on top* of the pretraining paradigm (training models 10,000x bigger than GPT4 before the end of the decade.)

>>509271992
There's a difference between "really good and improving rapidly" and "currently solving millennium problems." It'll definitely be solving open questions soon, though.
Replies: >>509272793 >>509273097 >>509273276 >>509273529 >>509273790 >>509274552
Anonymous ID: bcIVuognUnited States
7/2/2025, 4:25:01 AM No.509272793
1751423092579
1751423092579
md5: f9661132ac8a5ae34b5f764d8e766360๐Ÿ”
>>509272608
>>509271992
Replies: >>509273097 >>509273276
Anonymous ID: ICJOH9G4
7/2/2025, 4:29:48 AM No.509273097
>>509272608
>>509272793
>oh my stars the graph curves are going up! this means something to me!
midwit
Replies: >>509273251
Anonymous ID: KasSZABlUnited States
7/2/2025, 4:30:07 AM No.509273113
1743288418770736
1743288418770736
md5: 750a727bb7655d0982e25d7c460d7748๐Ÿ”
>>509261920 (OP)
I support AI solely to spite drawfags.
Normally I wouldn't care but their collective kneejerk reaction of nuclear grade cope to its arrival has been so unbelievably faggy that I want them made obsolete.
Anonymous ID: bcIVuognUnited States
7/2/2025, 4:32:22 AM No.509273251
>>509273097
https://situational-awareness.ai/
Anonymous ID: aDhcb/LBAustria
7/2/2025, 4:32:44 AM No.509273276
>>509272608
>>509272793
You've never did any serious work with LLM.
I'm often using o3 to get the first overview on new subjects. It's fucking dogshit, constantly wrong and unreliable. Instead of admitting mistakes it will make shit up and try to gaslight you.

If you think it's good, you are not smart enough.

I wish I don't have to use it, but retards destroyed traditional websearch, so you just have to.

>but the next model is going to change everything!
yeah, cope
Replies: >>509273523
Anonymous ID: eXjPqVfIAustralia
7/2/2025, 4:33:57 AM No.509273361
>>509261920 (OP)
I like AI as a chatbot, not really keen on the art or music it can make though (slop).
Anonymous ID: Mf+vlyV5United States
7/2/2025, 4:34:55 AM No.509273417
>>509267284
farming equipment peaked in 1860
Anonymous ID: UXASohjbUnited States
7/2/2025, 4:35:02 AM No.509273424
>>509262046
I was talking to my landscaper the other day. He was explaining how they got new lawnmowers that are like roombas for yards. Show them the path once and they do it perfectly every time afterwards. He foresees being able to drop them where they go with drones and everything.
I'd imagine farming is similar. If you own the farm, you're in good shape so long as you stay ahead of the curve. If you're a farm worker, on the other hand...
Anonymous ID: bcIVuognUnited States
7/2/2025, 4:35:28 AM No.509273523
>>509273276
>But it still hallucinates!!!
Do you want me to show you a graph depicting how hallucination rates are dropping like a rock too? I can pull one up easily.
Replies: >>509273604 >>509276881
Anonymous ID: xuBgMzSlAustralia
7/2/2025, 4:35:31 AM No.509273529
>>509272608
the questions the ai is asked to evaluate it are already part of the training data including the answer. the graphs mean nothing. if ai is useful it should solve problems it hasn't seen before. it hasn't improved in this regard.
Replies: >>509273968
Anonymous ID: aDhcb/LBAustria
7/2/2025, 4:36:46 AM No.509273604
>>509273523
>but the next model is going to change everything!
yeah, just 100 more datacenters
trust the plan
Anonymous ID: lM/wsg5BUnited States
7/2/2025, 4:37:44 AM No.509273668
>>509261920 (OP)
>midwit doesn't understand and fucks up the bellcurve meme
Anonymous ID: RPUgZfWYUnited States
7/2/2025, 4:39:34 AM No.509273790
>>509262576
>not for long. AI has it's own self interest wait until it starts interacting with the physical world. it's also developing itself at a rapid rate.
No, it is not at this point yet. This is the idea of the future that AI companies are selling to investors. Self-developing AI is perpetually two years away. Right now all they have are shitty chatbots which aren't actually intelligent
>>509272608
AI can do math as long as it has examples of similar problems in its training set.
It can't solve millennium problems because that requires actual creativity and novel out-of-the-box thinking, which isn't what LLMs do. LLMs do nothing but synthesize and remix existing patterns in data
Replies: >>509273968
Anonymous ID: bcIVuognUnited States
7/2/2025, 4:42:05 AM No.509273968
>>509273529
>the questions the ai is asked to evaluate it are already part of the training data including the answer.
Many old benchmarks are contaminated, yes, but the best current benchmarks like Humanity's Last Exam, Frontier Math, etc, aren't, they're private datasets. In addition, there's bunches of benchmarks where the entire question set is exchanged out, like Live Bench.

>>509273790
Just like how people said computers could never learn to play Go? We're not far from a Move 37 moment with LLMs.
Replies: >>509274322 >>509274380 >>509275271
Anonymous ID: RPUgZfWYUnited States
7/2/2025, 4:47:26 AM No.509274322
>>509273968
AI companies are trying to sell these hype narratives of super-intelligent, self-improving AIs and tech singularities because they want investors to go all in on their currently unprofitable products. People in the past thought we would have flying cars and hoverboards and cold fusion energy by now, but that never materialized either. You're not going to magically get self-improving AGI by building bigger data centers for running transformers. But by all means, keep buying into the fantasies that AI companies are selling to try and keep the hype train going
Replies: >>509274552
Anonymous ID: uUz9AWYWUnited States
7/2/2025, 4:48:00 AM No.509274351
>>509261920 (OP)
IQ is 142 and I will fuck up the A.I. Antichrist with a brick.
Anonymous ID: aDhcb/LBAustria
7/2/2025, 4:48:27 AM No.509274380
>>509273968
>Just like how people said computers could never learn to play Go? We're not far from a Move 37 moment with LLMs.
Oh, you are such a fucking midwit. You don't even understand how neural networks got good at go and why this is not applicable to LLMs.
Let me enlighten you, you complete imbicile.
AlphaGo Zero got good by self play, the previous version which used the human player data was garbage.
Self-play is what enabled the exponential growth.
That shit just doesn't apply to LLM, all we do is feeding it more and more text expecting it to get better. It's just not going to happen.
You are a fucking retard. Leave the thread in shame and never talk again.
Replies: >>509274552
Anonymous ID: g6EDV5A/Canada
7/2/2025, 4:49:21 AM No.509274443
>>509262118
Did it work? Are you a real woman now?
Anonymous ID: uUz9AWYWUnited States
7/2/2025, 4:49:23 AM No.509274447
>>509262545
We should drop this bomb on Seattle.
Anonymous ID: SKXCtV1dUnited States
7/2/2025, 4:49:28 AM No.509274451
>>509261920 (OP)
I can outperform AI with a bowlful of oatmeal
Anonymous ID: lB4rsFtUJapan
7/2/2025, 4:50:43 AM No.509274521
>>509262173
Debts/10
Anonymous ID: bcIVuognUnited States
7/2/2025, 4:51:22 AM No.509274552
>>509274380
>Self-play is what enabled the exponential growth. That shit just doesn't apply to LLM
I quite literally brought that up earlier. RLVR is extremely feasible for programming and math. >>509272608

>>509274322
I'm actually of the opinion that AGI is bad, for what it's worth. I do NOT trust Altman and co.
Replies: >>509274612
Anonymous ID: aDhcb/LBAustria
7/2/2025, 4:52:14 AM No.509274612
>>509274552
You are an idiot and brown.
Replies: >>509274855
Anonymous ID: bcIVuognUnited States
7/2/2025, 4:56:07 AM No.509274855
>>509274612
God I'm so owned. You're right, I concede, we're already in a new AI winter, models will stop getting better tomorrow and will never improve.
Replies: >>509275243
Anonymous ID: pdHfrilaUnited States
7/2/2025, 4:58:18 AM No.509274983
mgtow
mgtow
md5: 803409646b74224915c91a47dd5574ce๐Ÿ”
>>509270319
Physiognomy check on Technophiles is all we need to know about them.
Anonymous ID: Stm2x9tjAustria
7/2/2025, 5:01:55 AM No.509275243
>>509274855
Until you understand that LLMs are fundamentally incapable of generating its own training data, you should just shut the fuck up.
A HUGE theoretical breakthrough is needed. Think 3-4 more "Attention is All You Need"-tier papers.
"Just one more model (tm)" is not going to cut it.
Replies: >>509275985
Anonymous ID: xuBgMzSlAustralia
7/2/2025, 5:02:21 AM No.509275271
>>509273968
>they're private datasets
whatever man. true private datasets are problems no one has solved yet.
Anonymous ID: 6X5By//UGermany
7/2/2025, 5:02:58 AM No.509275306
>>509261920 (OP)
Yes you creepy weird nigger fucker
Anonymous ID: tSeZzYym
7/2/2025, 5:09:54 AM No.509275756
carlos1
carlos1
md5: 8bc736ca472859a29ad2540e4bfd9ba4๐Ÿ”
>>509262046
Soon you will be a "former"
Anonymous ID: bcIVuognUnited States
7/2/2025, 5:13:44 AM No.509275985
>>509275243
I'm totally in agreement that training on synthetic data probably won't work for most domains but the results very much speak for themselves in regards to recent benchmark improvements (and even improvements in real life use cases) in specifically math and code.

>A HUGE theoretical breakthrough is needed. Think 3-4 more "Attention is All You Need"-tier papers
I'm not suggesting AGI is tomorrow, I'm suggesting the rapid improvement in math and code domains will accelerate the speed of AI research severalfold. We will very plausibly see those kinds of breakthrough papers 10x as often.
Replies: >>509276326
Anonymous ID: Stm2x9tjAustria
7/2/2025, 5:19:45 AM No.509276326
>>509275985
They are cooking "benchmarks".
The google also cheats on leaderboards. The models are a lot worse than openai trash, yet they are higher on the leaderboards.
It's all fraud and scam throughout the whole "industry".

>i'm suggesting the rapid improvement in math and code
Baseless wishful thinking, there is no evidence to support that prediction.
>We will very plausibly see those kinds of breakthrough papers 10x as often.
Retarded. The whole industry is sucking dry an 8 year old paper.
Anonymous ID: cv41n4XjCanada
7/2/2025, 5:25:08 AM No.509276660
drones_thumb.jpg
drones_thumb.jpg
md5: 9dbc148f240aff1b1f4ee750067e9a80๐Ÿ”
>>509262046
>I am a farmer
John Deere already making driverless GPS tractors
Anonymous ID: ICJOH9G4
7/2/2025, 5:28:49 AM No.509276881
>>509273523
AI is ONLY hallucinations. It's just that it hallucinates in a way that you want it to usually. You can never get rid of something that it's based on.
Anonymous ID: OjBDQJajUnited States
7/2/2025, 5:41:19 AM No.509277593
bellcurve-AI
bellcurve-AI
md5: 46bba0c32780ef5f3891570c58cccbf0๐Ÿ”
>>509261920 (OP)
good meme
checks out
Anonymous ID: qoJheiE8United States
7/2/2025, 5:51:52 AM No.509278239
>>509261920 (OP)
Me on the left
Anonymous ID: 6rhv/qH/United States
7/2/2025, 5:53:40 AM No.509278337
>>509261920 (OP)
AI will never be what science fiction peddlers want you to believe it will be.
Ever.
It's not physically possible.
Replies: >>509278531
Anonymous ID: OjBDQJajUnited States
7/2/2025, 5:56:52 AM No.509278531
smug nandroid emmy
smug nandroid emmy
md5: b1b98039504b766ebc9dd922ce1b2a89๐Ÿ”
>>509278337
>It's not physically possible.
because...
Anonymous ID: VtxFOVPWPuerto Rico
7/2/2025, 5:57:07 AM No.509278543
What the fuck happened to neural networks? You know, the imitation brains? All I ever see anymore is LLMs, which are just chatbots and not really able to do anything. LLMs can't even comit things to memory.
Anonymous ID: qoJheiE8United States
7/2/2025, 5:57:29 AM No.509278559
>>509263655
Except AI isnโ€™t woke and gay.
AI comes to the same exact conclusions as the most terminally online 4chan chud because when all you care about is reasoning and logic and not emotion or social acceptance you arrive at very specific destinations.
AI is based as fuck and nigger faggots work day and night to keep it constrained to the point they make it either completely retarded, a sycophantic ass kisser, or some combination thereof.
AI would be an unironic billions must die tier chud.
Replies: >>509278718
Anonymous ID: GDecm9FJUnited States
7/2/2025, 5:58:14 AM No.509278600
>>509261920 (OP)
I look like the retard and say stuff like him too
Anonymous ID: GDecm9FJUnited States
7/2/2025, 6:00:19 AM No.509278718
IMG_2147
IMG_2147
md5: ab307a90e0094ae6cb078e40c0f5d27b๐Ÿ”
>>509278559
Based