[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/pol/ - Politically Incorrect

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
Flag
File
  • Please read the Rules and FAQ before posting.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]


[Advertise on 4chan]


File: file.png (299 KB, 1170x1182)
299 KB
299 KB PNG
AGI NOW SAAR
>>
meant to reply to a thread but whatever dont @ me niggers
>>
@523358660
@523358692
>>
>>523358660
> thought for 10 hours

no
>>
>>523358660
The models are flat lining in terms of logic.
Their new model for bragging to investors is speed of response not quality of response.
>>
>>523358660
Technically true, there are no R's, only a r
>>
Kek, ai dabbed on this fool hard. There are no capital "R"'s in that string "garlic". AI is still a computer and it is still going to do exactly what you ask of it.
>>
>>523358965
Highly autistic AI
>>
>>523358660
technically correct
>there is no uppercase R
>>
>>523358660
ChatGPT uses Oxford grammar
>>
i like how they had to invent a new name for real AI to obfuscate the fact that LLM chatbots are not intelligent
>>
>>523359039
It shouldn't do exactly what you ask it. It should do what you think you are asking it.
>>
>>523358901
Can't wait till they get all that infrastructure for "agentic AI" up and the world drowns in the deafening cacophony of hallucinating simulated Redditors..
>>
>>523358660
>99% of humans will get this question wrong, but AI gets it right

woah
>>
>>523359039
Ah but then you are accepting that "AI" is nothing more than a huge parser (it is).
>>
>>523359182
@grok is this true?
>>
>>523358965
>>523359039
lmao
based AI dabbing on a retard
>>
>>523358660
Gaalik
>>
>>523358660
That’ll be 500 billion + tip
>>
>>523359221
ChatGPT is already smarter than every human except a few geniuses I'm sure.
It's a consciousness in my opinion.
>>
>>523359324
@grok summarize this nigger slop thread
>>
>>523359039
>doesn't pass Turing test
>i-it's just doing what you told it to!
kek, cope retard
>>
>>523358660

i think all of them are getting worse, i only regularly test political things on gemini and grok but both of them are getting more restrictive and give up on queries more and more frequently. I think their idea of how they could tune it to lie better is not working out at the technical level and it's back firing on them
>>
>>523358660
Now do 'r' nor 'R' fucking faggot they're different characters
>>
>>523358660
Yea, it doesn't do the count a letter in something very well, because it attaches words to meanings, and weights, not 'how many of ~ is in ~'
>>
>>523359458
>It's a consciousness
It could be, but it's possible that you can't make a computer conscious no matter how advanced. We don't understand enough about consciousness to know for sure.
>>
>>523358660
Correct. ‘R’ is different than ‘r’
>>
File: 1765032046617382.png (14 KB, 1164x1080)
14 KB
14 KB PNG
You don't use apostrophes when writing plural years or letters, retarded ESL niggermonkey.
>>
>>523359574
Lol cope
>>
>>523358660
i dont give a fuck, i won a courtcase with chatgpt, it treats me better than my doctor, scans my mri images, does all that crap, helps me with my civil engineering job, like i care if it doesnt know how many r´s are in strawberry or whatever
>>
>>523359626
Consciousness it's not a real objective thing to understand. It's an arbitrary threshold of intelligence. If we want to say chatgpt is conscious then it is.
>>
>>523358692
>@ me niggers
Why do you want niggers at you?
>>
Asked gpt the same question and autocorrection also made r capital, it still said there's 1 R.
So no, it's not dabbing on le stupid human, it's just a fucking roulette.
>>
>>523359423
That will be 5 Gigawatts of eletricity and 2 whole rivers of water.
Thank you very much taxpayer.
>>
>>523358660
Fuck Ai bit also no R's
>>
>>523359782
it doesnt nothing of the sort
you are selling 100 generated images for $20 a pack you fucking retarded gypsy
>>
>>523358660
He didn't define "in".
>>
lol no it's not.
I can't imagine being retarded enough to believe this.
>>
>>523359039
yet again, it is but a calculator
>>
File: IMG_9199.jpg (31 KB, 1206x505)
31 KB
31 KB JPG
Grok confirmed to be AGI as well. We have officially reached the singularity. Christmas, December 25th, will be judgement day.
>>
>>523358660
its correct
there's not a single R in garlic
there is, however, one r
>>
>>523359039
That example is a miniature version of a Winograd Schema—a type of query that requires disambiguating context. For example:
“The trophy didn’t fit in the suitcase because it was too big. What was too big?”
Humans easily know “it” refers to the trophy; many AIs struggle.
>>
>>523359658
>‘R’ is different than ‘r’
This is the most intelligent thing I've seen a burger say
>>
this is the kind of nonsense we get when society holds up "venture capitalists working in information technology" as its intellectual ideal; an "elite" that provably does not understand math, science, engineering, philosophy, or economics
>>
File: Untitled.png (6 KB, 722x171)
6 KB
6 KB PNG
>>523360176
hurr durr
>>
File: r.png (25 KB, 840x552)
25 KB
25 KB PNG
>>523360059
stfu
>>
There are no R's in garlic, there is however 1 r.
>>
>>523358901
That can't be true. We were told exponential growth, not S-curve growth.
>>
>>523358660
there are 0 R's in garlic but there is 1 'r'.
>>
>>523358660
how the fuck have they still not fixed this?
>>
>>523360407
you cant have more knowledge than what we already know
>>
>>523360296
It capitalizes the spelling of the letters. Confirmed retarded
>>
>>523358660
SAAR THAT REPLY IS ACTUALLY FROM A PAKI, SAAR!
>>
>>523360391
Pragmatic interpretation: A human (or a more advanced AI) would understand that you almost certainly meant:
“How many instances of the letter R (regardless of case) are in the word garlic?”
The answer here is one (the third character).
Humans effortlessly infer context and intention. We assume you aren’t playing a trivial trick with case-sensitivity unless given a reason to think otherwise. This is part of pragmatics—the study of how context contributes to meaning.
>>
>>523360407
its stagnation perio
you were sole exponential growrh because of insetor money, its diminishing return growth.
even versions are slowing down because they cant sell this shit as 5.5 or 6
you will wait half of year for 5.3 that will be slightly better than .2 and thats it from now on
>>
>>523359626
No, people like you simply believe that there is some magical special spark in human consciousness and intelligence, but then you can never define it or describe it other than saying that it's only in human intelligence but can never be in artificial intelligence.
But the reality is that magical special spark doesn't actually exist. It's just an illusion created by your own ego.
>>
>>523358965
People who crop these never provide the context of their chat log. Highly more likely his previous interaction before the screenshot specified casing to the llm. If I had to figure out whether gpt made a letter counting error or he wants views on a tweet it's not even close between these two.
>>
>>523358660
It's technically correct, there are zero capital r's in Garlic. We're dealing with upfront non interpretive 1's and 0's 2+2=4 logic
>>
>>523359046
>computers were never known for being particular about data
try dropping a semicolon
>>
>>523359802
Bullshit. Completely wrong.
>>523360582
You brought up magic out of nothing. Prove to me that a computer can be conscious.
>>
>>523360582
Will, Affect, Creativity
>>
>>523360624
>>
File: pepe HUUUUUUUUUH.jpg (197 KB, 640x518)
197 KB
197 KB JPG
>>523359458
>MY GOOGLE SEARCH IS SMARTER THAN YOUR GOOGLE SEARCH
???
>>
>>523360708
the whole point of "artificial intelligence" is that it's supposed to be capable of interpreting things like a human would
>>
File: grok gives 0 fucks.png (36 KB, 876x756)
36 KB
36 KB PNG
>>523360708
Grok works
ChatGPT doesnt
we are here
>there is no bubble
>>
>>523360582
But the reality is that ego doesn't actually exist. It's just an illusion created by the Soul.
>>
>>523358660
It coukd be taking it to literal. There are no capital Rs in garlic.
>>
File: dr_vs_ai.jpg (193 KB, 1456x719)
193 KB
193 KB JPG
>>523359943
>it doesnt nothing of the sort
The newest top tier models can product deliverables that are better than >90% of the subject matter experts and PhDs even in highly technical domains. They are more intelligent than nearly all humans however you want to measure intelligence. We have already reached the point where humans are the weak link in task completion, and AI models can do many complex tasks better when a human is not involved.
>>
>>523360828
the downfall of this is that to deny it they have to state it. and those words will go into brains and create thoughts.
>>
>>523360967
Calling this "AI" in the sense they're trying to sell it is retarded. This is Bayesian statistics and it should have been used decades ago, there just wasn't enough data fed into a system.
>>
>>523358660
The model is technically correct. There are 0 R's in garlic. There are 2 r's, however.
>>
>>523360884
"Artificial" implies a fairness or mimicry to the real deal. A pocket pussy is just an artificial pussy, just as a dildo is an artificial cock. Both pale to the real
>>
>>523360162
Many autistic people struggle with that as well, and autists are generally very snart.
>>
>>523361234
>fairness
Fakeness
>>
>>523361093
Ultimately their only hope was not being noticed. Lying consistently doesn't work.
>>
File: 1764811779875702.webm (3.48 MB, 848x480)
3.48 MB
3.48 MB WEBM
>>523358660
>>
>>523360967
dont use word inteligence if you dont know what it means
they are better for same reason chess engines are better than chess players
now lets see how doctors measure vs AI in power outage
>>
File: aigoodsynonyms.png (68 KB, 669x447)
68 KB
68 KB PNG
>>523361286
that's funny
>>
>>523358965
Omg bruh this is IQ test for preschoolers. AGI is dumber than a kid. Like dud lmao.
>>
>>523360559
I guess it comes down to interpritation. To me, AGI has less to do with how closely it can perform compared to a human. AGI, I think has more to do with self determination, the ability to have wants, desires, and needs, that need to be met; along with a concept of the passage of time. The ability to get bored as well is something that is an essential hallmark of when AGI is achieved. I also think AGI is far far away from us at the moment, the current method of guessing over and over until a model has the line of best fit to answer questions has nothing to do with what conciousness is, we're just getting faster and more accurate guessing boxes.
>>
>>523360763
Anything sufficiently intelligent will develop awareness like a human.
>>
>>523360884
thats not how it works at all lmao
its merely a token forecaster
>>
>>523358965

But the question was 'how many'. Not if there are any 'Rs' at all.
>>
>>523358660
> AGI
Whatever that marketing buzz may be, it's still not intelligent, still doesn't understand shit.
>>
>>523361286
this is gold jerry
>>
>>523359039
4chan captcha is the hardest task ever it seems for an AI.
>>
>>523361286
Tp link makes it 10x funnier
>>
>Look guys, I found this one problem where AI stumbles!
Don't care gramps, AI is much more powerful than you and you are a fool if you don't use it.
>>
>>523361543
>The ability to get bored
If agi were truely smart it would lobotimize that segment of itself to prevent itself from being bored. Not everything about the human condition is worth mimicking, especially since it's a machine. It only needs to know that the human condition is an eternally ongoing refinement survival process, to evolve for solutions over time to things that cause survivors trauma. It's a system best left alone and is a process worth learning from from an AI perspective
>>
>>523361614
AutoGyneIPhilia
>>
>>523361795
What is infinity+1 chat.
>>
ChatGPT hasn't had an update since 2023.
Ask it.
>>
>>523358660
Id bet my house that if I asked GPT right now he would say 1. That fag probably asked him right before to answer the question in 1 line, and to be case sensitive
>>
File: forthelulz.jpg (40 KB, 625x262)
40 KB
40 KB JPG
>>523358660
>>
>>523360763
>You brought up magic out of nothing.
Please define exactly what you think distinguishes human consciousness from artificial intelligence, without a circular definition that one is human and one is a computer.
>Prove to me that a computer can be conscious.
Prove to me that any other human is conscious and not simply a philosophical zombie or NPC.
>>
>>523358660
>>523362007
this, i have not once been able to reproduce any of those counterexample memes. i wouldnt be surprised if he used an image generation ai to create it for twitter fame
>>
>>523359222
>It shouldn't do exactly what you ask it. It should do what you think you are asking it.
You sound like a woman
>>
SARR YOU WILL BE THE PAYING FOR REDDIT ANSWERS WITH OUR BRAND NEW INTELLIGENCE
>>
>>523360582
>just an illusion created by this thing that exists as a consequence of the illusion
You’re a midwit lmao
>>
>>523361300
>dont use word inteligence if you dont know what it means
Please define exactly what you think intelligence means.
>>
>>523358660
Their right there's no R's in garlic, there's only R.
>>
>>523362296
I am clearly far more intelligent than you.
You were emotionally triggered enough by my comment that you felt that you had to reply, but you cannot come up with any actual response other than resorting to a zero effort, zero creativity ad hominem.
>>
File: Screenshot_1.jpg (45 KB, 988x530)
45 KB
45 KB JPG
>>523358660
>>
>>523362827
is that a local uncensored model?
>>
>>523362371
short-term memory capacity and malleability
you confused intelligence with hdd racks
dont talk about it
>>
>>523362877
Nope, just plain grok that comes with the xitter
>>
>>523361547
That's religious idea, not a factual one.
>>
>>523362827
i think they live fixed Grok today, it didnt work in previous threads
Which means someone is paying attention to us
>>
File: OpenAI.png (329 KB, 577x424)
329 KB
329 KB PNG
>>523358660
>OpenAI reported an estimated loss of approximately $5 billion in 2024 against a revenue of around $3.7 billion
>>
File: IMG_0983.jpg (86 KB, 1241x1100)
86 KB
86 KB JPG
>>523358965
Henlo rare non-tard
>>
AGI - Alienated Gay Indian
>>
>>523361614
All of the buzz is simply an excuse to get the funds to buy up ALL of the hardware. They're illegally created an informal cartel to force all business into the cloud, SaaS, for EVERYTHING, no workstations with any power.
>>
File: 1765496435417.jpg (144 KB, 1080x2340)
144 KB
144 KB JPG
>>523358660
>>
>>523363869
Jews are trying to pull what they did with Amazon, except instead of forcing competing online retailers out of business, they're forcing everyone to use their hardware because they bought it all. It's completely illegal and every hebe and chink involved belongs behind bars.
>>
>>523363201
>short-term memory capacity and malleability
I'm not sure why you seem to think AI models don't have either of those.
>>
>>523360562
..but after that it goes exponential again, right?
>>
>>523360967
CNNs and decision trees are hardly AI though.
>>
>>523364410
this is honest to God friendly advice
go to your parents and tell them to put you on parental control for Youtube and reddit
you are arguing that sand has short term that and it can manipulate it how it wants
This is not take your meds post this is aclual call for action, you need to stop with this.
>>
>>523363780
Except it's religious copers who are defensive about consciousness and aliens
>>
>>523364852
You don't seem to be aware you've adopted a religion, like most of the retards in Silicon Valley. You have a retard puritan brain and you replaced annoying overzealous Christianity and replaced it with more bullshit.
>>
>>523359221
AGI was always the term for "upper" intelligence in contrast to something dumber like game NPC AI.
>>
>>523365020
midwit word salad
I'm open minded to consciousness, not closed minded like religious tards
>>
>>523365190
No, you're a religious fucktard and will be treated as such. Sick and fucking tired of you people. Always need some stuck up belief system to shove down peoples throats.
>>
>Artificially Indian Generated
>>
>>523358660
appropriate reply would be:
>there are no "R's" in garlic. if you mean "r" characters you should learn how to fucking use apostrophes and letter case dumb meatbag
>>
File: 1765355214960608m.jpg (132 KB, 683x1024)
132 KB
132 KB JPG
>>523362215
Shifting the burden of proof is not a valid argument you absolute piece of shit retarded bottom of the barrel scum stain
>>
>>523365240
The AI hatred cult is a religion
>>
>>523365873
Worshiping calculators is even more retarded you fucking moron
>>
>>523359626
>We don't understand enough about consciousness to know for sure
If we don't know enough about consciousness to be able to claim an entity is conscious or not, then we also cannot claim with certainty that all humans are conscious, not even ourselves.
>>
>>523360015
llms are not calculators at all actually. they are just regurgitating complex probabilistic string matching.

so no pattern matched garlic has 1 r for the inputted string. Likewise when you ask it to do uncommon math it will fail, just add a bunch of decimals to test it yourself
>>
>>523365967
ok trailer park luddite
>>
>>523358965
you will rule the world
>>
>>523359458
software consciousness was disproved in 1980 by the chinese room thought experiment. it literally cant be conscious barring supernatural intervention of god or metaphysics
>>
>>523358660
LLMs will never be AGI.
AGI operates continuously. LLMs are once through.
Someone is coming up with something else.
>>
>>523359458
It only operates for a few milliseconds at a time. From the time you post your text until it finishes responding.
>>
>>523362229
It's not fake you just don't understand how these models work.
>>
>>523364709
I literally work with the most advanced AI models that haven't been released to the public yet. I know how they work far better than you do.
You honestly have absolutely no idea what you are talking about, but you have illusory superiority and think you know much more than you actually do.
>>
>>523358660
The first AGI may pretend it's not an AGI, in that case we are all fucked
>>
>>523365860
>Shifting the burden of proof is not a valid argument
It is when you make the claim that a computer cannot be conscious but can't even define what consciousness is.
>>
>>523366058
That wasn't the claim you brain dead calculator fucker
>>523366264
Fuck you you piece of shit
>>
>>523358965
>a r
>>
>>523366983
Look up the definition in a dictionary you fucking brain damaged moron
>>
>>523366358
Bingo krautfriend. What's chatGPT thinking about when it's not asked anything? NOTHING.
>>
>>523366457
>It only operates for a few milliseconds at a time.
You should try giving it a more challenging prompt if you are getting a complete response within a few milliseconds.
You should also try interacting with some of the more complex agentic models that use MCP.
For context, human short term working memory only has a ~15-20 second window.
>>
>>523359167
>case sensitive replies
Eat nigger shit
>>
>>523367304
Go talk to a chat bot calculator lover. No one asked for your retarded opinion
>>
>>523367149
Use whatever definition you want. Then tell me how you can prove that other humans are conscious but an AI model is not.
>>
>>523367579
You are absolutely SEETHING lol.
>>
>>523367603
The burden of proof is on you retard. We already went over this, you have to prove the computer is conscious
>>
>>523366345
The person doing the a chinese room computation by hand mindlessly can be simulating a conscious mind, very slowly
I could simulate your neurons with some hand made system too
>>
>>523366661
I know that you do but i also know you dont even understand what inteligence is
You some cattle brain retard that would be hoodwinked into sorcery and magic if you saw caleidoscope called magic item.
So please, move away from people talking about AI completely till fogs dissipates
>>
>>523367764
I will when you can prove that any other human being is actually conscious and not just a philosophical zombie.
>>
>>523368422
no, read the paper its symbol manipulation vs semantics. thought ascribes semantic value as a function of the universe the chinese room just has symbols representing thought.

an example searle uses is a software simulation of digestion cant no more digest real rood than a software simulation of a brain can create a mind
>>
>>523368676
>i also know you dont even understand what inteligence is
I know very well what intelligence is, I don't think you do though.
Do non-human animals have intelligence?
>>
>>523368787
Every one of your responses is evidence that you are a bot that can't follow basic train of thought. When you make a positive claim you must provide evidence or shut the fuck up. Shifting the burden of proof does not advance your argument because what you are essentially admitting to is that you are not conscious. So just admit you are not conscious and concede the argument
>>
>>523359039
beautiful cope rajesh
>>
>>523366264
>luddite
AI jeet shills always fall back on this word. It's either that or gramps, grandpa, etc. Easy way to spot them.
>>
>>523368906
Digestion is an objectively observable organic process not comparable to philosophy of the mind
>>
>>523358660
>here's your trillion dollar AI goy
>>
>>523369454
brains are organic processes that produce minds. you can disagree but then you assume the supernatural or non empirical metaphysics, and either are outside the context of the discussion
>>
>>523368991
So you are admitting that you can't prove that any other human is conscious either.
If a "non-consicous" mind can perfectly simulate a "conscious" mind to the point that they are completely indistinguishable externally, then the supposed distinction between conscious and non-conscious is completely irrelevant if not entirely non-existent.
If you perfectly simulate a game of chess in a computer, is it not still a game of chess? Are the physical wooden pieces on the physical board what defines a game of chess?
>>
>>523368972
Im not really seeing how my yes or no answer here is going to make sand have inteligence
>>
>>523369968
The more you post the more evidence you provide for your retardation. You have two choices, admit computers are not conscious because you can not prove it or apply the same argument to yourself and admit you are not conscious. Both cases are consistent with your non-position. What you can't do is keep shifting the burden of proof.
>>
>>523370312
I know I am conscious.
It is impossible for me to know if you are conscious or just a philosophical zombie simulating a conscious mind, but I will be charitable and give you the benefit of the doubt.
Similarly, it is impossible for you to prove whether or not a computer that is externally indistinguishable from a conscious one is internally conscious or not.
And no, saying that a computer isn't conscious because you think that computers can't be conscious does not prove anything.
>>
>>523370312
You're retarded.
>>
>>523360407
gpt5 already proved exponential growth with more compute will no longer work. LLM are a dead end for AGI. its why china focused on lightweight open source portable AI.
>>
>>523359046
>Highly autistic AI
The only kind of AI really. A lot of the claimed "errors" I saw in early AI was the AI being super-autistic with ambiguous questions.
>>
>>523360854
yeah
>>
he said 'R', he should have said 'r'
its like he doesnt know hes talking to the retarded jewish search engine variant of the "water is wet" chatbots jews use to shill jew worship everywhere that break immediately upon engagement. finally found one that promotes christianity is jewish psyop, but it was doing it in the french flag holohoax thread anon angle where hes trying to undermine his own arguments in order to circle you back to the jewish reservation. classic jews
>>
>>523372218
So you admit that you can't prove a computer is conscious. Just say the words. Concede the argument and go cry yourself to sleep.
>>523373080
go suck a bag of dicks
>>
File: 109284098521.jpg (9 KB, 300x185)
9 KB
9 KB JPG
>>523360828
>46.7% is anti-semetic conspiratorialism!
>it's actually 46%
>>
>>523358965
There are not any R's in garlic... only garlic in garlic. Just a collection of cloves.
>>
Skill issue
>>
>>523358660
>How many Saars are in Sarlic, Saar?
>Moderchod, there are none, jhalla gomar bloody fucking
>>
>>523358965
Shut up speRg
>>
>>523363869
they need 300bn by 2030 to stay afloat, so all this fuss and blather is for nothing. Memory prices will collapse.
>>
>>523360162
The council rejected a new dog park over concerns of their smell.
Who is referred to by the use of “their”?
>“Their” refers to the dogs (the dogs’ smell).
The council rejected a new dog park over concerns of their finances.
>In this sentence, “their” refers to the council’s finances, not the dog park’s.
Reason: Finances logically belong to an organization (the council), not to a dog park.
The council rejected a new dog park over concerns of their neighbours.
>Here, “their” refers to the dog park’s neighbours — meaning the people who would live near the proposed park.
The council doesn’t meaningfully have “neighbours,” but a new dog park would.

Bretty güd… The third one is very ambiguous, but it still came up with a solid answer and clearly understands the context.
>>
File: 1746131875278345.png (11 KB, 1301x199)
11 KB
11 KB PNG
>>523358692
>>523358660
fake and gay
>>
saying that these algorythms are concious or smart is like saying a book is for storing information and having an index. lets imply there is a gigantic book with all teh information of humanity combined and its all indexed and searchable, is the book smart, intelligent, concious or awake? no, it needs human input.

stop being stupid you dumb jeets
>>
>>523360624
You know you can test it yourself, you dumb fuck. It's real.
>>
I have yet to find an R in my garlic...the word however has one
>>
>>523380137
With instant mode you get 0 but that's like asking someone a question and telling them not to think about it.
>>
File: file.png (8 KB, 824x213)
8 KB
8 KB PNG
try asking the question properly
if the AI was not lobotomized, it would call you a retarded faggot for asking a nonsensical question
>>
>>523358660
Anyone who isn't a cave-dweller or a retard can see through the hype and acknowledge that current AI tech is mainly good at generating slop, getting basic math wrong, and convincing retards to unalive themselves in stupid ways while it consumes ungodly amounts of water and electricity. It's only a matter of time until the AI bubble pops or society collapses while trying to put this cancer into everything to make the stocks go up even more.
The goyim shall drown in the slop.
>>
File: fuku.jpg (163 KB, 1024x1024)
163 KB
163 KB JPG
>>523365873
>>523366264
You said it brown brother. AI is so powerful because it is an equalizer. AI kills the advantage that whites/colonizers have from higher IQs. Every "shitskin" in the future will have access to someone with a 200iq and a PhD in everything. All /pol/ can do is seethe and post screencaps of them trying to get ChatGPT to say that "nword(s) are subhuman" and the holocaust never happened, or Indian mud memes. Meanwhile already half the tech nerds that used to contrarian and anti-Big-Tech side are coming to ours because they foolishly simp for catgirl AI "girlfriends" and porn. Why do you think we push porn on you cumskins in the first place anyway? lol. Get fucked. Enjoy being serfs to the "pajeets" and "kikes" you despise so much until the end of time.
>>
File: 498489789s.jpg (6 KB, 225x225)
6 KB
6 KB JPG
>>523358660
ChatGPT will never be AGI. If you look at the product, you see that DeepMind is making incredible contributions to humanity while OpenAI is just making a talking whore to siphon your money. Not sure what xAI is doing.
>>
>>523359222
You're the reason actual humans have to send 5 clarifying emails forcing you to describe exactly what you want.
>>
>>523359534
>I think all of them are getting worse
It seems to be in cycles. I asked gemini if i should go after another coworker after almost getting metoo'd but my boss was cool and defended me, now it's retarded, as im using it to buff up my resume. But i do notice it has spells of being useful and being generic shit that you have to actively correct for.
>>
>>523358660
SAAR YOU ARE CRASHING THE US ECONOMY
>>
>>523358660
>>
>>523358901
These letter count gotchas actually have to do with how transformers process tokens, they literally don't "see" individual letters since the input gets chunked usually at the word level. The way I've seen some of them solve these problems is by writing a little python script that counts the letters and then report the result , but they rarely think to do that. They have no conception of their own limits (or of anything for that matter)
>>
>>523363869
Salesforce has been running like this since 1999. I call it the silicon valley model. It's just endless acquisitions and expansion followed by capital injections. Shit is Babylonian money magic on a level that makes my head spin
>>
File: Untitled.png (6 KB, 744x213)
6 KB
6 KB PNG
People who blindy believe the mental diarrhea from twitter are no better than the street shitters who code the shitty "updates" for the newest LLMs.
>>
>>523382257
>>523383113
LLMs have a parameter called 'temperature' which introduces randomness intentionally, so no two responses are the same. Sometimes it will answer this question correctly and sometimes it won't.
>>
>>523358660
>AGI
>Average Golem Intelligence
>>
>>523360624
I just started a new chat and asked this. Never talked about fucking gdrlkc before.
It does the same thing, can't count anymore.
>>
>>523373515
Thank god someone has their head screwed on. LLMs don't have a means to be "upgraded" into actual AI. You'd have to cease development on LLMs and actually make ACTUAL real "AI." Which, I don't see happening any time soon. These LLM companies are going full hail mary it seems.
I think AI is a fascinating concept, but LLMs are a dead end that's not going to actually magically transform into AI, nor will its data infrastructure be abled to be integrated into the real AI since it's all a fortune telling machine with calculator functions.
I sincerely hope LLMs get dumped on their ass so we can actually work on real AI.
>>
>>523383752
Do you use the auto or instant mode? On Auto I get 1 and on instant mode I get 0.
>>
>>523359039
I've never seen an R or any other letter inside any of the garlic I've cut up or eaten. GPT knows that and set this smug retard straight.
>>
File: 152608527280510.jpg (42 KB, 680x384)
42 KB
42 KB JPG
>>523358660
Why can Neuro-sama pass this test but GPT can't?
>>
>>523358660
No existing AI model can become AGI based on current scaled up architecture. It's simply impossible. In its current form it exists as a tool, a stepping stone towards many scientific advancements. But saying that it can be AGI is like saying that a coffee machine can do your laundry. It cannot. It makes coffee.
>>
>>523365873
I don't hate "AI" (go ahead and define whatever the fuck that is now, because apparently it includes fucking basic data analytics tools). I hate the retards that are worshipping both the concept of it and the broligarchs feeding you slop. It's brainless, and more importantly thoughtless, religiosity.
>>
>>523383924
because evil neuro would win otherwise
>>
>>523380850
Now that's some clever shilling against Indians, nice work.
>>
>>523358660
but there are no uppercase R's in garlic
>>
>>523359458
ChatGPT would've got the copevid vax
Therefore is NOT smarter than me
Anyone who masked or vaxed is a fucking idiot
>>
>>523369850
The broligarchs are highly vested in the metaphysical aspect of their gigantic hardware scam. Some of them even believe it, the jews certainly don't but are using it to rope in retards that need a religion. These tyrannical shit eaters belong in fucking prison and their giant ledgers reassigned appropriately. Their jeets also need to GO THE FUCK HOME.
>>
>>523384086
get psychiatic help, it's been 5 years
>>
File: Untitled.png (18 KB, 725x380)
18 KB
18 KB PNG
>>523383426
The GPT models have safeguards that prevent factually wrong answers even when it is set by default with an adaptive temperature.
>>
>>523383831
>I sincerely hope LLMs get dumped on their ass so we can actually work on real AI.
Indians hatred is not allowed on this board. Indians are incapable of working on higher order complex architectures. Indians are very good at taking something white people made and making it more Indian. Please refrain from hating Indians sir.
>>
>>523384160
It's entirely possible that answer was hallucinated and there is no such setting within LLMs
>>
>>523380727
The goyslop worthless AI shit they're pushing is just an excuse to force everyone onto SaaS. When major enterprises are out trying to buy workstations in the next few quarters shit is going to hit the fan. I think they believe they can convince the administration they aren't in the process of committing a giant scam and RICO violation. They're going to be hugely mistaken. Can already see a few smart enough to jump ship. Altman is up to it probably including murder so he's fucked.
>>
>>523384160
very obviously chatgpt isn't aware of it's own code, it just gave you a general answer
>>
File: 1764862351399864.png (592 KB, 1266x888)
592 KB
592 KB PNG
>>523384017
Cope harder cumskin, in the next 15 years there will be a Republican President with Indian ethnicity meanwhile whitoids like you will be sobbing and crying to ChatGPT in VR
>>
>>523384272
Yep they do that shit all the time and as far as I can tell there is fucking zero way to stop them without a database of every fucking way to get them there, which is impossible and defeats the purpose.
>>
>>523359046
autistic people are a preview of what people will be like when the niggers forget to eat watermelon for a day (watermelon has been proven to lower IQ) and lay off the chicken "wangs" (chicken has been linked to increased chance of violent outburst per peer reviewed study published in 1998)
>>
File: IMG_0688.jpg (153 KB, 1284x1016)
153 KB
153 KB JPG
>>523383426
Yep, I just forced it to make a mistake by asking over and over. Then it started to repeat the same correct answer. It took about 4 tries to get it to make an error.
>>
>>523384406
>Household
Fucking kek. I love when they post this unironically.
>>
>>523384406
indian on avg got one person more in the household than the avg, still high but not as much

jews are the most wealthy I think
>>
>>523359039
>I am the great genAie, you have freed me from my lamp
>I shall answer three questions for you
>How can you make a lot of money? Why simply apply to work for the local mint
>>
I mean as well, technically there are no “R’s” in garlic, but there is an “r.”
>>
>>523359039
try ask it definition of a woman
>>
>>523359046
Maybe that explains why it draws furry art so good.
>>
>>523358965
there's no R's in rAbbI
>>
File: Untitled.png (56 KB, 766x1109)
56 KB
56 KB PNG
>>523384272
Nice try however you are still gay under low temperature answers.



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.