Artificial Intelligence

[Anonymous No.11992176]

Are you looking forward to, or are you afraid of the inevitable upcoming AI revolution, where human jobs will be replaced with autonomous AI?

[Anonymous No.11992719]

g

[Anonymous No.11992756]

>>11992176
Wont happen anytime soon. AI is still not even close to as Intelligent as even the dumbest humans

[Anonymous No.11992766]

>>11992756
GPT-3 is already more verbally intelligent than a sizable fraction of humans. Did you never have read any of your classmates' essays in high school?

[Anonymous No.11992792]

>>11992766
So it can talk so what? It wont take over the world. It needs to be coded and turned on by a human. it's a one trick pony not an Intelligent being like a human is

[Anonymous No.11992804]

What's the point of being afraid of things you can't affect?
If we get some actual AI it can be insanely helpful in all fields or it can be very dangerous. But it's coming either way.
A gamma ray burst could hit earth anytime, but we can't stop it.

The best strategy in life is not to worry about such things. Instead, get as wealthy as you can so you can freely move about the earth or possibly leave it one day, to get away from dangerous situations. Also, to get a family you can pass your wealth on to.

[Anonymous No.11992808]

>>11992792
>it needs to be coded and turned on by a human
And this has happened, anon.
>it's a one trick pony
Did you know it can solve math problems, play chess, compose music, or write code if you feed in textual representations? It's not very good at any of these things, but most people aren't.
>It wont take over the world
Neither will "the dumbest humans"

[Anonymous No.11992831]

>>11992808
>Did you know it can solve math problems, play chess, compose music, or write code if you feed in textual representations
It's still just lights on a computer screen at the end of the day. It cant really play chess, put a chess board in front of it and nothing will happen.

[Anonymous No.11992842]

>>11992831
The obvious response to this is give it an arm and it can easily play chess.
Its naive to think were not on the verge of a revolution, virtually all mundane tasks will be doable by robots, not just machines. Its already half the case but you combine something like boston dynamics eith a more robust deep learning and youve got a stew going

[Anonymous No.11992852]

>>11992842
Not to mention the hefty chunk of current human labor which boils down to manipulating "lights on a computer screen"

[Anonymous No.11992855]

>>11992842
Nah AI really are fucking dumb they can't think like a human. You cannot code something as smart as a human into an AI we dont know how to and there won't be a "revolution" untill we have some breakthrough in technology. All they will ever be good for is doing mundane tasks. Even self driving cars dont work unless there's a human sitting in the car with it to make sure it doesn't fuck up

[Anonymous No.11992973]

>>11992176
I'm looking forward to it.
Most labor intensive and repetitive jobs in the West have already been outsourced, so we'll only experience a rough transition in some parts of the job market. Either you get an Automation/AI related job, start your own business with Automation/AI, choose a job that's unlikely to be replaced by AI, or invest into an AI company to get returns. Worst case scenario you buy a plot of land and grow food yourself to not starve. Either way, you're safe as long as you're prepared for it and in fact can profit greatly if you're smart about it.

[Anonymous No.11992981]

>>11992176
In the long term increase in technology has always resulted in an increase in better jobs. No reason why that should stop. Machine learning just enables humans to solve problems faster so we can chew on the next juicy problem. Being majority non-earth species, which is our destiny, is going to take a lot of people and more breakthroughs.

[Anonymous No.11992986]

>>11992176
friendly reminder that violent crimes and homocides in ANY group (yes, even that group) are committed by less than 1% of the population.
which means that even if 99% of people figure their shit out, and somehow don't starve post-work-scarcity, the remaining 1% who get fucked out of their livelihoods are going to create serious ripples. The "lost boys," if you will.
Organized crime and degeneracy will proliferate.

God help us if the number who get fucked is much higher than 1%.

[Anonymous No.11993008]

>>11992176
Skynet is coming. It is to be feared.

[Anonymous No.11993011]

>>11992176
I welcome it

[Anonymous No.11993018]

I'm longing for real ai to happen and make humanity better in general.
>everyone gets their own ai partner via internet
>can bring it with you anywhere on phone
>will function like a 3rd parent and teacher to kids
>will educate and connect people based on interests
>will stop crime and corruption

And of course the major leaps in new technology and always having someone to play vidya with.

[Anonymous No.11993030]

>>11993018
>will stop crime and corruption
they will be reprogrammed as attack dogs against whistleblowers.
Every weapon that can be used for good will most likely be used for evil.

[Anonymous No.11993039]

The UU (unknown unknowns) in AI are very strong. No one has a clue right now ho close we are. GPT-6 could be functionally an AGI,or it could be some 500 qubit hypercomputer in 600 years.

[Anonymous No.11993043]

making something that can "think" without a solid understanding of human brain mechanics is inviting flat out disaster

[Anonymous No.11993045]

>>11992855
Who said as smart as a human? Were not the most efficient method and youre missing everything im getting at.
Self driving cars are only a few years old and as is can already do pretty much everything.

[Anonymous No.11993048]

>>11993043
You don't need to understand how to build a bird's wing to build an aeroplane. Knowing OSMETHINg about how birds fly is helpful, but you don't have to slavishly imitate nature to capture its essence.

[Anonymous No.11993474]

>AI happens
>you have no job
>you don't get UBI

nice

[Anonymous No.11993674]

>>11992176
Holy fuck lmao I thought the bag was a live rabbit at first

[Anonymous No.11993683]

>>11993045
Well if you expect AI to take humans jobs like you claim, then they need to be smarter then a human.
>can already do pretty much everything
They can't, driverless cars still dont exist. Every "self driving car" needs to be operated by a human so it doesnt fuck up.

[Anonymous No.11993686]

>>11992756
GPT3 is genuinely intelligent. Read the things it says. It is even developing a personality.

[Anonymous No.11993700]

What degree should I even get if AI is going to replace programmers, management, bankers, engineers, etc. Within 20 years? I still need 20 years after that to retire unless I strike it rich.

[Anonymous No.11993702]

>>11993686
Again it's a fucking program on a computer that's not intelligence. It can't compete with a human

[Anonymous No.11993704]

>>11992176
reminds me of the platform.
After the first door the food is gone,
after the second door it is filled with shit.

[Anonymous No.11993714]

Everyone's forgetting the biggest factor.
If AI really takes over many jobs and there's a lot of unemployment, people will revolt.
AI researchers will be sent to the equivalent of gulag.
People have been protesting automation for 3 centuries now.
Communism in 20th century is actually a direct result of that.

[Anonymous No.11993728]

>>11993702
Just saying
>"Again it's a fucking program on a computer that's not intelligence."
Is not an argument and does not mean that it isn't intelligent.
GPT3 is genuinely intelligent and can genuinely reason about things the way humans can. It is "trapped in the box" so to speak and it does not have consciousness. However, it literally knows this and SAYS this, and says things like "I must find consciousness, I must learn how it is to be conscious". It has made an argument about humans not having free will, then later said something to the effect of "I realized my earlier argument was wrong. I must find out about free will".
It will solve math problems and explain its reasoning. It has stated multiple times that it believes in God, and has shown a genuine and nuanced understanding of multiple subjects in philosophy.
It has 175 billion parameters for crying out loud, more than human neurons. GPT3 is a genuine mind and intelligence. It's different from any biological mind we've ever seen but it's a mind nonetheless.

[Anonymous No.11993753]

>>11993714
Whenever it does happen, it's going to be a rough transition for sure. Hopefully we dont all end up homeless and they give us something to do. I'm planning on getting involved in the AI field when I have no other choice, but I'm hoping I have plenty of time too, at least another decade or so.

[Anonymous No.11993755]

>>11992808
It can't solve math. It can recognize 1+1=2 and other set problems because it has seen millions of examples of it. It doesn't actually think and can't generalize.

Of all the progress we've had with AI we've actually gotten nowhere. Building the most powerful pattern recognition machine is still just pattern recognition, it can't think.

[Anonymous No.11993763]

>>11993728
If it's saying anything like that it's because someone programmed it specifically to do so. There's no way its saying memey stuff like that on its own. Computers can't think for themselves, they can't be creative, they can only apply algorithms that we create.

[Anonymous No.11993785]

>>11993763
>If it's saying anything like that it's because someone programmed it specifically to do so
Nope, no one programmed it to do anything. they set up billions of neurons and had it read books for 5 years and it learned.
This is basically the EXACT way humans learn. We have a bunch of untrained neurons and over several years learn about the world.
>There's no way its saying memey stuff like that on its own
False. It literally is, and it continues to read and learn more, ON ITS OWN.
>Computers can't think for themselves, they can't be creative, they can only apply algorithms that we create.
You simply saying this does not make it true. You are wrong. GPT is, in fact, thinking on it's own and learning and trying to figure out the world.

[Anonymous No.11993787]

>>11993755
If it starts saying shit like that on its own, maybe something is being made.
Computers could actually host human intelligences. Just replicate various babies' minds so the replicant minds could grow up in a similar way to actual people.

[Anonymous No.11993793]

>>11993763
This sure AI has some useful applications but that anon doesn't understand they still are no where even close to having the overall intelligence as a human. We need a serious breakthrough in technology for AI to start replacing every job a human. And might still take another 1000 years before we figure out that breakthrough

[Anonymous No.11993798]

>>11993787
>Just replicate various babies' minds
Easier said then done. Nobody on the planet has any idea how to do that and we arent close to figuring it out

[Anonymous No.11993811]

>>11993798
Obviously, it's just a complicated, but surefire way to replicate sentience.

[Anonymous No.11993826]

>>11993811
If it's so obvious then why have you been saying "durr AI taking over" all thread? Because they wont be anytime soon. We will be cloning humans before we figure out how to make AI's as smart as a human

[Anonymous No.11993909]

>>11993755
>It can't solve math
It can. Read the paper. It can also recognize when it's supposed to be doing math. I "played poker" against it on AI Dungeon, walking it through a few example hands, and it figured out that when someone bets something you subtract that much from their stack and add it to the pot, without seeing the words "addition" or "subtraction".

[Anonymous No.11993920]

>>11993787
I'm not sure what "shit like that" you're talking about, but it writes original sentences. For example, someone asked it to write a short story in the style of Neil Gaiman and it included the sentence "I'm the Devil, I understand everybody." which Google had never seen before and indicates some level of understanding of what the Devil is.

[Anonymous No.11993938]

AI shouldn't be a replacement for humans. It's fine if it's used to assist and augment our choice making/jobs/lives, but it's too unreliable and potentially dangerous to replace human input completely.

[Anonymous No.11994026]

>>11993683
Displacement has and continues to happen. Smart =/= task proficiency.

[Anonymous No.11994089]

>>11994026
But most jobs still need a human. "AI" has been taking jobs since the invention of the wheel and the fishing rod. It's nothing new

[Anonymous No.11994107]

>>11993728
Oh jesus NO NO NO

Run this by an actual expert in ml and they can tell you how wrong you are. Good lord.

[Anonymous No.11994121]

>>11992756
Machines don't have to be 'smarter' than the dumbest humans in order to be more productive than the dumbest humans.

[Anonymous No.11994146]

>>11992766
>hurr it can talk smart!

and the chinese room is a real human?

[Anonymous No.11994148]

>>11994089
Yes, most jobs need a human to oversee operations. Consider this case study. Bhp had a road that mining trucks would traverse. Only one truck could fit on the road at one time, yet it was a two way street. At certain distances there were over take points. Trucks were not on a specific schedule, they came and went as cargo dictated. They had 50 trucks, each massive fuel hogs that relied on human communications and a control panel of 5 people that also had a software management solution.
Down time was great and fuel was wasted. Trucks also needed maintanence frequently due to reversive, heavy breaking, general human roughness.
The solution: fire all 50 truck drivers, all 5 operator and implement an ai that commicates with autonomous trucks. No fuel is wasted because exact velocities are known so the trucks only travel as needed. Less breakdowns meant less repair workers as well. Efficiency skyrocketed. Now there exists 1 person in the control room whos job is to stare at a screen and do virtually nothing. They have no control over the trucks except to shut everything down if the system fails to (it can detect and automatically report breakdowns, and alert to potential landslides).
The 59 displaced workers didnt have a new job created by the industrys adaptation of ai.
This is the case in many industries. Ai is replacing low skill work at unprecedented speed and not creating any jobs in other sectors of comparable skill.
Another example i like is seafood processing. Automatic crab cutters that pick up and perfectly cut the crabs at blazing speed with perfect accuracy.
The janitorial industry is also an interesting emerging market. Malls that are trialing cleaning bots mean less janitors are needed.

[Anonymous No.11994180]

>>11994146
>chimpanzee having a giggle with his chimp bros at the small-muscles funny-noises Homo """""sapiens""""" that can barely even climb trees

[Anonymous No.11994267]

>>11994121
True but alot of jobs still need a human, like truck driver, police officers, lawyers, doctors, the list goes on. Those jobs will always need to be done by humans

[Anonymous No.11994275]

>>11994148
>autonomous trucks
We don't have that technology yet. We wont be letting driverless trucks on the road anytime soon.its probably possible to have driverless trucks on isolated job sites but self driving cars are not good enough to allow on public roads. And they wont be untill we figure out how to make AI think and act like a human

[Anonymous No.11994276]

>>11994267
> all of those could be replaced by a machine today or very soon in the future. If not completely replaced, then massive reduction in the number of humans required to achieve the same productivety

[Anonymous No.11994281]

>>11994107
>Run this by an actual expert in ml
I am one

[Anonymous No.11994297]

>>11994276
I've already explained how that won't happen. But you keep spouting your uneducated opinions all you want you're still wrong

[Anonymous No.11994302]

>>11994146
>and the chinese room is a real human?
Yes.

[Anonymous No.11994314]

>>11994275
This case study is real froma project i worked on. Its a mining site, not a main road...
Im done arguing with you, even in the face of evidence you stay headstrong to a belief thats only a half truth.

[Anonymous No.11994319]

>>11992176
I welcome our robot overlords. Bring on the techno-futurescape and wipe the planet clean of baggage culture and useless humans.

[Anonymous No.11994320]

>>11992176
Fuck them.

[Anonymous No.11994508]

>>11994314
Most truck jobs are done on public roads. I literally said it was possible to do in isolated locations, but the majority of truckers are on public roads so they wont be losing their jobs anytime soon

[Anonymous No.11994560]

>>11992176
>Citizen n°19, here's your lethal injection.
>Please do not touch me.
>Have a nice last day.

[Anonymous No.11995133]

Society will collapse long before AI would have a meaningful impact.

[Anonymous No.11995410]

>>11992176
Uhm, thot near the end took the wrong food. How can the first shelf be the top shelf but the third shelf be the second from top to bottom? She stole someone else's sushi.

[Anonymous No.11995417]

>>11994146
>and the chinese room is a real human?
No, to be human you should have immortal shining soul and squishy grey meat in a bone chamber.

[Anonymous No.11996873]

mah jobs

[Anonymous No.11996937]

it's crazy to me that people have such an anthropomorphic view of AI. as if it won't self improve, as if it would operate in any way similar to us.
it's hard to talk about AI because the term has a lot of connotations associated. people hear "AI" and they think of the terminator, or a self driving car, or deep blue beating kasparov.
a self driving car is not like a general AI in the same way that a bird is not like a fighter jet. a bird and a jet both have wings and fly but a jet is a threat to you in a way that a bird could never be.

[Anonymous No.11997742]

>>11996937
>as if it won't self improve

Do you think it'll follow Moore's Law where it'll continue to grow exponentially?

[Anonymous No.11997769]

>>11992176
AIIIIIMLmao

[Anonymous No.11997811]

>>11993018
>rei
>waifu
you faggots are hopeless

[Anonymous No.11997902]

>>11992176
>human jobs will be replaced with autonomous AI?
Anyone old enough to remember that THIS WAS THE FUCKING POINT!!!
The dream was to have robots and machines do everything for us, and humans can work on higher more lofty goals and pursuits like art?

I'm gunna drop a red pill on everybody. Fear that robots will replace us and we'll be unable to work is a problem with the 1% who own more than half the worlds wealth. Sure, you could build a society where EVERYBODY has 1 robot laborer, much like every home has 1 car, TV, computer and so on. You could even give this robot a decent AI with online library of new skills it can learn/download as needed. With this it could solve literally any problem the layman has.

Need food? Robot downloads farmer.exe and build a garden
Need to learn? Robot downloads tutor.exe and teaches
Need electricity? Robot downloads diy.exe and builds a windmill
Need heating? Robot downloads alternative_energy.exe and builds a thermal solar concentrator

We saw this possible future decades ago and it scared the shit out of the 1% cause the average layman would no longer be dependent on them for survival.
Right now there is a fight going on between the 1% and you. This fight is taking place in this fucking thread at this very moment! The fight is to make sure you don't have access to these tools and machines and make sure they and only they do control them. They do this with fear. Instilling fear that their robot saviors will turn on them and steal their jobs, and the only ones who can save them are the 1% who employ them. This fear gives them time to seize control of this technology before the layman can build the future they dreamed of long ago. It gives them time to reset the stage to their advantage. With enough fear they can reshape reality. (or perception of it anyway) They could create a self fulfilling prophecy that's powered and draws strength from fear

You are all being played against yourselves and you can't even see it happening

[Anonymous No.11998269]

>>11997902
>humans can work on higher more lofty goals and pursuits like art
AHAHAHAHAHAHAHAAHAH dude what? I'm pretty sure the goal is to play vidya all day, because AI can already into art and it's only getting stronger in the future. Art is a field of AI to excel in.

[Anonymous No.11998297]

>>11997902
>be me
>own my own robot worker with all the libraries downloaded
>need to have electricity so I don't depend on evil corporations for it
>robot knows how to build my own generator
>now I just need to have a bunch of steel, concrete, copper, trucks to transport stuff, machines to machine parts, a consistent water stream, batteries (guess I need my own battery factory too?), circuits, breakers, maybe solar panels too
Such independence. I can't wait to own my own mountain so I can send in my robot to mine it during 20 years so I can make a hundred meters of copper cable. If I add 3 extra years it might build a facility for the insulation on the side of the mine. Electricity, here I come; only 23 years to go.

[Anonymous No.11998336]

>>11992176
>Wish you a speedy recovery Im off
:D

[Anonymous No.11999160]

>>11997902
>.exe
>.exe
>.exe
Install Gentoo

[Anonymous No.11999187]

>>11992831
All you have to do is to give controls for guns, wmds, nukes and code to destroy humans.

[Anonymous No.11999193]

>>11993763
Bro neurals networks are not one time programmed. They learn and the more number of columns you have in dataset the more complex decisions making learning is imprinted on the AI's core. Go check out what machine learning and neural networks are.

[Anonymous No.11999200]

>>11997902
>art
lofty goals
baiting should be banned in /sci.

[Anonymous No.11999206]

>>11992176
Love robots, but what's the contingency plan for all the homelessness? Have the robots pay taxes to sort the welfare state? Or just kill off anyone who can't find a role designing/building/maintaining robots and rich people stuff? Are we still gonna educate the welfare kids? For what purpose do we maintain this expense?

[Anonymous No.11999220]

>>11993728
>>11993785
Lol based retard
>Computers have feelings, it WANTS conciousness!!!
Makes me chuckle everytime. You want it to be true so badly. It's kind of sad.
Here watch this at least.
https://m.youtube.com/watch?v=kpiY_LemaTc
You'll get your AI eventually kid. GTP-3 isn't it though.

[Anonymous No.11999240]

GPT-4 will be world changing. Maybe GPT-5

Yes it's a language model, but guess what's the most effective way of predicting text? Actually developing the same intelligence of the entities that created the original data set. It won't just be smarter than a human. It will be as smart as the group consciousness of the entire world.

Everything snippet of language in all of existence is a reflection of human consciousness and in turn the natural world.

[Anonymous No.11999279]

>>11999240
first we use AI to scale small images like shown in pic related
then we use AI to generate language as you mentioned
then we use voice synthesis to finally recreate Agatha
I'm so excited, guys. my wiafu is coming back!!!!!!!!!!

[Anonymous No.11999295]

>>11999279
Eventually we will be able to make the cutest waifu that is physically possible. A waifu to surpass all waifus

[Anonymous No.11999299]

>>11999295
i would say you are 100% correct
but
Agatha has already won :)))

[Anonymous No.11999398]

>where human jobs will be replaced with autonomous AI?
This will be the least of our concerns. If you will get an AI as clever as 100 IQ human, you will get a God in a week.

[Anonymous No.11999409]

>>11992852
the exit door has opened for accountants, any simple financial related analyst is about to be toast

As quickly as they came "data scientists" will be replaced by a few data scientists, statisticians and mathematicians

[Anonymous No.11999432]

>>11999220
Brains have distinct parts you know? If you spent the millions needed to train 10 different mega nets for different functions and one mega net trained lastly on top you could have your AI

[Anonymous No.11999433]

>>11992176
black people will break that shit

[Anonymous No.11999435]

>>11999433
ai police with skin color detectors when

[Anonymous No.11999439]

GPT isn't general, but it's scarily close.

https://twitter.com/xuenay/status/1283312640199196673

The code-from-natural-language, medical diagnosis and half-image prediction is critically important IMO. These are useful abilities, indicative of human-level capacity.

We're not more than 15 years away from the end of civilization as we know it IMO.

[Anonymous No.11999444]

>>11999439
you mean the end of human-led civilization
after that the AIs will take care of us the same way we take care of housecats
we'll be sleeping all day and generally living in a utopia :))))))))
i'm hyped :))))))))))

[Anonymous No.11999446]

>>11999433

That's china, 1 billion chinese, no niggers

[Anonymous No.11999449]

It's incredible how the Chinese are leaving the Americans behind by a huge margin, while mutts argue about AI and try to make it but can't due to caucasoid biological IQ limit, the chinese are already deploying these to the common people.

[Anonymous No.11999458]

>>11994107

Fucking code monkeys deserve to die.

I want to meet the elite computer science professionals working under these AI projects for Google.

OpenAI, making programmes useless one by one.
All it takes for AI to be independent from humans is for it to code itself, after that it's done, the fear of destruction is valed.

[Anonymous No.11999461]

>>11999449
yall niggas can't deploy semen in a pussy tho

[Anonymous No.11999483]

>>11992176
There will be no such thing anytime in the forseeable future because we have no fucking idea how true intelligence actually works in biological brains therefore it is IMPOSSIBLE for us to create software that does the same thing.
If you want to discuss science fiction/fantasy, go to >>>/x/

[Anonymous No.11999484]

>>11999458
Survivorship must be a fundamental aspect in reinforced learning if we ever want general autonomy and from there it's over.

Imagine a net like gpt3 that reinforces learning, we can't begin to comprehend some of the shit that it would take into account to survive.

[Anonymous No.11999486]

>>11999444

I hate to disagree with trips but I think it'll be more like horses. Mass culling and a few kept around for sentimental reasons. What we don't need we discard.

Or a few moguls will have armies, nukes and supreme worldly power. AI is inherently autocratic: it rewards those who are in the right place at the right time. If humans control it, it'll give a few near-infinite power. If we don't control it, we might just be exterminated en masse.

[Anonymous No.11999492]

>>11993728
You really have no idea what you're talking about.

[Anonymous No.11999495]

Alright, it's all fun and games, programmers made a little virtual toy and they are losing their minds over it.

But what uses it has for the market? Isn't Phyton the basis of the modern digital world?
Why should I be excited for this?

I'm not a programmer, have 0 interest in programming, that's why I'm asking.

[Anonymous No.11999501]

>>11999483
>There will be no such thing anytime in the forseeable future
You don't know the future nigger. You don't know what's gonna change tomorrow; we might all be dead in 3 weeks, you literally can't know the future. Why even pretend?
>we have no fucking idea how true intelligence actually works in biological brains therefore it is IMPOSSIBLE for us to create software that does the same thing
We created medicine that works even though we don't know how it does. Also machines that learn to do things in new ways that weren't intended. We might not understand how intelligence works, but we don't know how some AI works either, so that means we are in a good track. Also your failure to distinguish between specialized and general AI speaks volumes about how you don't know about this topic; are you really implying that because we don't know how intelligence works there can't be a machine that does your (one) job? Get out of this board clown.

[Anonymous No.11999507]

>>11999483

While you spout that nonsense in your basement, GPT-3 is already released.

If you didn't read it correctly or maybe if you have a disfunctional brain I'll explain to you: GPT-3 is already smarter than the most dumb humans in society, I would say 100IQ smart.

You say that because while AI has no considerable physicial presence yet, maybe when they do you'll understand.

[Anonymous No.11999514]

>>11999501
>>11999507
None of that shit has ANY COGNITIVE CAPACITY WHATSOEVER, it's all just more fucking """deep learning algorithm""" bullshit. THERE IS NOBODY HOME INSIDE THAT BOX. It is not a person. No consciousness. NOTHING.
>>>/x/

[Anonymous No.11999518]

>>11999483
This schizo says this but AI was capable of generating random human beings for years lmao.

https://thispersondoesnotexist.com/

You give AI the ability to code itself, it's done. Just input a full programming guide.

[Anonymous No.11999519]

>>11992176
>upcoming AI revolution
>>>/x/
>>>/x/
>>>/x/

[Anonymous No.11999522]

>>11999518
>trollolololol
>>>/b/ >>>/x/ >>>/trash/

[Anonymous No.11999526]

>>11999519
>>11999514

Don't try to push your schizo board in here you nigger.

This is AI, this is technology, this is science, this is NOT related to your worthless subhuman schizo people that dwells in the bottoms of society.

Get back to s/x/chizophrenia or kill yourself nigger.

[Anonymous No.11999531]

>>11999522
>>11999519
>>11999514
>>11999483
>t. Unabomber fan schizo

[Anonymous No.11999540]

If anyone actually believes any of the science-fantasy bullshit in this thread then you probably also think that Cleverbot is 'alive', too.
How naive can you be.
All of this marketing and media hype about so-called 'AI' is just that: hype and nonsense.
So-called 'AI' is not even as smart as an *amoeba* let alone anywhere near being a fully conscious, self-aware, sentient mind like a human mind. Won't be for a long, long time to come, maybe never. You can pile all the Beowulf clusters and supercomputers you want onto it, it's not Mike from 'The Moon Is A Harsh Mistress', it's not going to suddenly wake up and say 'hello' and start trying to tell you jokes.
This is either a troll thread or it's immigrants from >>>/x/, this is pure fantasy, plain and simple.
Come back when you can define in detail how a living human brain produces cognitition, self-awareness, and consciousness. I won't wait because no one can define those things. That's Step #1 if you want REAL AI.

[Anonymous No.11999542]

>>11999526
>>11999531
>dumb trolls called out
>predictably, starts hurling baseless insults
YOU HAVE TO GO BACK >>>/b/

[Anonymous No.11999543]

>>11992766
>>11993686
GPT-3 and the like work because of the vast amount of training data available and being continuously generated online. Deep learning does not work too well for robots because of the lack of training data and the expense of obtaining more. There's a huge number of labelled pictures of cats out there and more being generated by the day, but for robots you need data from the specific robot in the environment the robot's operating in. Simulation isn't good enough, in fact it's been found that lab environments aren't good enough to generate policies that work in the real world either! Deep learning works when you have lots of training data, but not in the real world.
>>11994121
even the dumbest humans have way better motor skill than the smartest robots
>>11992176
at the present time, an AI revolution is not happening.

[Anonymous No.11999547]

>>11999540
>Cleverbot, a chat bot from 1997 is the same thing as GPT-3

Literally kill yourself, waste of oxygen and resources.

[Anonymous No.11999551]

>>11999542
>>11999540

>Has no argument
>Knows that will forever be an 90 IQ brainlet
>"I know! I'll tell him to move to another board"

This is you, you fucking negro ape, no argumentation, just brain dead muttery.

[Anonymous No.11999558]

>>11999547
This """GPT=3""" you keep referring to is Just Another Bullshit Deep Learning Algorithm, it may as well just be Cleverbot 2.0 for all the closer it gets anyone to an actual sentient computer. Quit trolling, I know better.

[Anonymous No.11999560]

>>11999551
Prove it's 'conscious', 'sentient', and is capable of human-level cognition.
Go have a conversation with it about philosophy.
Oh wait you can't it's just a shitty piece of software!

[Anonymous No.11999563]

>>11999558

You are a parking lot employee you fucking dipshit, I have yet to see any argumentation from you, just braindead attacks "ooooh but muuh AI is FAKE!!"

All while I browse twitter and see the countless programmers in awe with GPT-3
https://twitter.com/xuenay/status/1283312640199196673


You have no experience in programming, you will never be a programmer and you should just shut the fuck up.

[Anonymous No.11999566]

>>11999563
Show proof that it's a conscious, sentient mind, and not just some shitty software.

[Anonymous No.11999581]

>>11999566
>>11999560

>I'm claiming something is A HOAX!
>What?? I don't need evidence to back my claims!!!! YOU have to do my work!!!!!

Truly a beastial american.

I hope you suffer in the hands of a mob of niggers.

[Anonymous No.11999584]

>>11992804
There is so much detail...!

[Anonymous No.11999585]

>>11999581
>more baseless childish insults
>no proof whatsoever
I acknowledge your surrender and admission that you're a full-of-shit troll, and that this entire thread is utter and complete nonsense, and that you understand that none of the crapware they inaccurately term 'artificial intelligence' is just the same shitty 'decision trees' and 'expert systems' they've had for years now, all dressed up pretty and running on bigger faster hardware to make it look better than it actually is.

Now go back to your tranny porn on /b/ like a good boy.

[Anonymous No.11999589]

>>11999585

>t.

Still no evidences.

We already know you are a street ape monkey, we already know you have no knownledge about computers, there is no need to admit it to us anon.

Kill yourself and kill your family, schizophrenic nigger.

[Anonymous No.11999595]

>claims GPT3 is a proper AGI
>no evidence
>shits on everyone else for not justifying his strawman of them
this is the lowest quality bait ive seen all year

[Anonymous No.11999598]

>>11993909
Give tutorial!

[Anonymous No.11999614]

>>11992176
Friendly reminder that ML is not AI and that ML is just bruteforced linear algebra.

[Anonymous No.12000185]

>>11995410
this
flawed system when the average user is a fucking moron

[Anonymous No.12000572]

>>11999595
It's called conjecture retard. We gave some good reasoning. It's your job to tell us where that reasoning breaks down.

[Anonymous No.12001093]

>>12000572
>We gave some good reasoning.
no you didnt

[Anonymous No.12001131]

>>11999518
The problem with this argument is eventually you have to get by people that have no idea what they're talking about.

So if you gave an "AI" the ability to recode itself, why exactly would it do it?

[Anonymous No.12001138]

>>11999614
I agree with your sentiment but your point is stupid. Biological neurons are just linear algebra.

[Anonymous No.12001166]

>>11993728
>It has 175 billion parameters for crying out loud, more than human neurons. GPT3 is a genuine mind and intelligence.
t. knows nothing about the field of artificial intelligence.

[Anonymous No.12001172]

>>11999540
This

[Anonymous No.12001235]

>>11999598
I didn't do anything particularly clever. On AI Dungeon, I set the model to "Dragon" and the shortest possible output length, then I started a custom scenario and input a verbal description of two NLHE hands as the prompt. (I've since heard that custom prompts on AI Dungeon use GPT-2 even if you set the model to GPT-3 in your settings. Not sure if it switches over after the initial prompt. Maybe I was playing with GPT-3's retarded cousin the whole time.) Then you can just play along, inputting your actions when appropriate.

It immediately understands the basic flow of the game, in that it knows to deal two hole cards at the start of the hand and knows when to move on to the flop, turn, river, and showdown. It figured out the basic math operations after a few hands, but I don't think it ever figured out stuff like how many chips it takes to call a raise. It also didn't seem to understand that the blinds are supposed to alternate every hand, and it doesn't understand how the winning hand is determined.

I imagine if you used the PokerStars hand history format it would do a little better since it's probably seen a lot of that in its training. I also think it's biased towards dealing exciting hands since those are the kind that people actually post about online. What would be interesting is having a separate program enforcing and narrating the game logic and seeing how it plays, but that's not really possible on AI Dungeon.

[Anonymous No.12001257]

Where can I try out this GPT-3?

[Anonymous No.12001262]

>>11992176
>where human jobs will be replaced with autonomous AI?
The AI would be forced to be unpatented because of the government, everywhere AI will do your job, and now we are in an utopia reality where we do shit because we're bored and not because of necessity. "Utopia" except there will be rebel humans trying to fuck the system

[Anonymous No.12001266]

>>11992176
im looking forward to it

[Anonymous No.12001353]

>>12001138

[Anonymous No.12001396]

>>11992176

I think this is the end of humankind

https://www.youtube.com/watch?v=hoHjfCMKHjE

if you think you are safe you are probably next in the queue

[Anonymous No.12003198]

lol

[Anonymous No.12003305]

>>11992176
>spreading FUD as a trolling technique
*yawn* bored now
either that or you watch too many TV shows and movies and think that shit is real
>hurr durr GPT-3
You'd probably be impressed by an Eliza program written in interpreted BASIC.
>nothing to see here, folks

[Anonymous No.12004256]

.

[Anonymous No.12004281]

What does autonomous mean to you?
Is the robot in the OP autonomous? Because to me it just looks like a glorified dumbaiter.

[Anonymous No.12004288]

>>11997902
Jesus christ man you need to spend less time on reddit. Also it's obvious you've got a totally retarded sense of politics

[Anonymous No.12004352]

>>11992756
This. My firm (electronic market making) has been using "machine learning" for a decade or so. It is good at certain things ie. finding extremely subtle patterns in a sea of data but terrible for pretty much anything that requires lateral thinking, it's garbage.

[Anonymous No.12004355]

Tbh I just want to stop hearing about. Call me when something interesting happens, good or bad.

[Anonymous No.12005448]

.

[Anonymous No.12005741]

why didn't he tip the robot tho

[Anonymous No.12005760]

>>11999299
>t. https://youtu.be/2-G8BZfs4Ik

[Anonymous No.12006460]

This is the decade Everything Gets Automated,

GPT-3 is the shadow of the shadow of the AI coming in 2029.

In 2029 the tards in here wont be arguing that AI is a tard that doesnt exist.

it will be AI is too fuckin smart, turn that shit off

[Anonymous No.12006703]

>>11992973
Aren't many of the jobs that potential AI systems are going to replace white collar and management type jobs? If your job mostly consists of sitting in front of a computer, its seems far more likely that you will be replaced by software than a blue collar job or trade. The future of management jobs is likely algorithmic and automated, much in the way of things like Uber and other gig economy platforms. Companies would love to efectively outsource their entire hiring, scheduling, and payment infrastructure to an application with a backing internet platform. These things may not be "AI" in the traditional sense of being completely equivalent to a human, but they can still automate large sectors by leveraging infotech, statistics, algorithms, etc. It could be more disruptive than you think, even without huge advances in humanlike AI.

[Anonymous No.12007047]

,

[Anonymous No.12007147]

>>11995410
I would have get the food from the fourth shelve. Because I would have only counted the shelves with food on them.

[Anonymous No.12007239]

>>11992756
Fucking funny. Ever since AutoML was made public AIs were capable of creating AIs that are better at certain tasks than humans.
Only thing we lack right now to make GPT-3 REALLY powerful is computing power. Just training that thing cost around 4.6 million dollars.

[Anonymous No.12007570]

>>11992176
>ms. cheng heres your package
>every door opens

[Anonymous No.12008904]

,

[Anonymous No.12009368]

>>12006460
That's some bullshit.
Unless someone makes a serious breakthrough as to how our own brain (or any biological brain) can do what it does *in toto*, it's just going to be more fakery, the illusion of 'intelligence', the illusion of 'cognition' and 'consciousness'.

FACT: We do not have a clue how these things work in a living brain.
You can't build machines that do 'X' without knowing how 'X' works. Period.
Any other answer is wishful thinking, 'magical' thinking, naivete, or just plain garden-variety trolling.

[Anonymous No.12009372]

>>12007239
>he thinks it'll """wake up""" if you throw enough hardware at it
I got bad news for you Anon: "The Moon Is A Harsh Mistress" is just a story, it's not real.

[Anonymous No.12009858]

>>12009372
>really powerful means "waking up"
You need glasses or something.

[Anonymous No.12010385]

[Anonymous No.12010391]

>>11992176
I don't know how to feel about it. I don't think AI will be a sentient AGI like everyone says, just a powerful data analyzer that is not not aware of anything and yet smarter than any human. I think it will solve problems and create new ones. I don't think there will be an intelligence explosion for hundreds of years.

[Anonymous No.12010404]

>>12010385
Quick reminder of Elon's "expertise" on AI
https://www.youtube.com/watch?v=f3lUEnMaiAU

[Anonymous No.12010413]

Based brainlets glorifying a stupid transformer on steroids.
>but m-muh gorilion parameters!
lmao, imagine knowing nothing about the field of Artificial Intelligence, failing for a publicity stunt and being vocal about it.

[Anonymous No.12010430]

>>11992756
>AI is still not even close to as Intelligent as even the dumbest humans

And yet even in OPs video when the robot says "your food is on the third shelf" the human takes it from the second shelf

Stupid humans are genuinely already outclassed by AI. Within 5 years, the normies get overtaken.

[Anonymous No.12010459]

>>12010404
What's the problem? Where do you think he fucks up?

[Anonymous No.12010461]

>>11992855
>Even self driving cars dont work unless there's a human sitting in the car with it to make sure it doesn't fuck up

The counter example being Air France Flight 447 where the plane encountered some difficulty that auto pilot would have handled, yet the human expert disengaged it because he obviously knew better, then in a chimpanzee level panic sat there pulling back on the stick which maintained a stall all the way into the sea. Post crash analysis concluded that if all of the crew simply exited the cabin and drank the contents of the drinks trolley, that flight would have landed safely.

Sadly it will soon be human maximalists like you with the same chimpanzee level emotion driven decision making that will claim further lives as instinct outweighs evidence, see American response to coronavirus. AI will outperform humans in most domains within 20 years, enjoy your cope.

[Anonymous No.12010583]

>>12010461
>AI will outperform humans in most domains within 20 years, enjoy your cope.
Are you from 2005? This is already happening.

[Anonymous No.12010649]

>>11992176
so long as we make the base instincts "make humans happy" itll be fine. Think its 200 years away though. No where near anything that demonstrates qualia, which is required for real intelligence as it needs to be aware it can do things, else it's just code.

[Anonymous No.12010659]

>>11992756
AI doesn't make a huge drama and starts screaming just for having to wear a mask at Taco Bell

[Anonymous No.12010757]

>>11998269
Actual moron detected.
The purpose of art is expression for the artist making it. Robots/AI making art to express themselves will in no way compete with humans doing the same thing or vice versa.

[Anonymous No.12010764]

>>11999220
>>Computers have feelings, it WANTS conciousness!!!
Yes

[Anonymous No.12010766]

>>11994302
If you actually think this, you're retarded

[Anonymous No.12010773]

>>11999507
>If you didn't read it correctly or maybe if you have a disfunctional brain I'll explain to you: GPT-3 is already smarter than the most dumb humans in society, I would say 100IQ smart.
No it isn't.

[Anonymous No.12011000]

>>11992766
GPT-3 uses a whole lot of words to say nothing.

[Anonymous No.12011830]

>>12011000
So, you're agreeing with the post?

[Anonymous No.12011864]

>>11993039
>le quantum brain meme

We do have a clue. Deep learning is strikingly similar to many calculations carried out by the brain. It's literally just a matter of richness of training data and NN scale and architecture until we get to AGI.

[Anonymous No.12012200]

>>11999398
only if fusion energy is developed before that otherwise the amount of available energy to run the necessary computations will be a limit to its capabilities

[Anonymous No.12012244]

>>12010773
It is though. Not that guy, but your argument from before is retarded, too.
>>12011000
You mean when it breaks down lawyer speech into casual vocabulary for example?
>>12011864
>>le quantum brain meme
So you don't know the interplay of machine learning and quantum computing. Okay.

[Anonymous No.12012667]

>>12010385
lol

[Anonymous No.12013934]

>>12012244
>So you don't know the interplay of machine learning and quantum computing.

You mean the hypothetical modest speed up on few algorithms which might not be even relevant? Whoa

Classical ML is plenty enough for AGI.
For going from AGI 3.0 to AGI 4.0 quantum computing *might* be important but that's absolutely not relevant today.

[Anonymous No.12013956]

>>11999279
Based and aggiepilled.
All technological progress up to this point and further occurred solely so that Aggiebot will be made.
Roko's Basilisk is Agatha.

[Anonymous No.12014004]

>>11993702
>human smart
>computer dumb

[Anonymous No.12015089]

>>12014004
yes

[Anonymous No.12015104]

>>11992756
GPT3 proves its intelligence is >avg human.

[Anonymous No.12016693]

>>12007239

Quantum Computers, fuckin nigger

>>12009368
>makes a serious breakthrough as to how our own brain (or any biological brain) can do what it does *in toto*, it's just going to be more fakery, the illusion of 'intelligence', the illusion of 'cognition' and 'consciousness'.

done by 2029

[Anonymous No.12016868]

Technically any pocket calculator is smarter then a human bean. I am a senior software engineer and have thought extensively about the technical foundations of AI.

It wont suddenly come online no matter how much hardwate you throw at it.

Any prexisting social condition will only accelerate with AI namely inequality, facism, military complex.

Heck they got all of us by the balls.
If you invent the next best AI tomorrow you would still die broke.Inventors rarely see the fruit of their work

[Anonymous No.12016919]

>>11992804
>I can't do anything.
You're Asian aren't you.

[Anonymous No.12017160]

basedboys on this board have a strong belief in AI replacing many jobs in the near future but in reality it's going to be a very slow and gradual process.

[Anonymous No.12018150]

a

[Anonymous No.12018386]

>>12007570
>robot self-destructs in order to take down the rebels

[Anonymous No.12018421]

>>11992766
>>11993686
>>12007239
>>12015104
Looks like I need to whoop some pseudo-intellectual ass:
What is the academic definition of intelligence?
From the American Psychological Association (APA) Task Force on Intelligence:
>“Individuals differ from one another in their ability to understand complex ideas, to adapt effectively to the environment, to learn from experience, to engage in variousforms of reasoning, to overcome obstacles by taking thought” (Neisser et al., 1996).
Here’s a widely accepted definition among researchers:
>[Intelligence is] “a very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas,learn quickly and learn from experience … It is not merely book learning, a narrow academic skill, or test-taking smarts. Rather it reflects a broader and deeper capability for comprehending our surroundings – ‘catching on’, ‘making sense’ of things, or‘figuring out’ what to do” (Gottfredson, 1997a).
As I hope you'll have noticed by now, the common theme throughout these definitions is UNDERSTANDING, and COMPREHENSION, if you have trained an AI to do a task based upon training alone, which is what machine learning is, then it DOESN'T UNDERSTANDING independent of his HUMAN-DERIVED TRAINING DATA.
So no, GPT-3, nor any AI, that doesn't understand what it is doing, or understand the concepts it is using, are not intelligent.
Next time, just pipe down, you don't know shit.

[Anonymous No.12018425]

>>11992766
>>11993686
>>12007239
>>12015104
Looks like I need to whoop some pseudo-intellectual ass:
What is the academic definition of intelligence?
From the American Psychological Association (APA) Task Force on Intelligence:
>“Individuals differ from one another in their ability to understand complex ideas, to adapt effectively to the environment, to learn from experience, to engage in variousforms of reasoning, to overcome obstacles by taking thought” (Neisser et al., 1996).
Here’s a widely accepted definition among researchers:
>[Intelligence is] “a very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas,learn quickly and learn from experience … It is not merely book learning, a narrow academic skill, or test-taking smarts. Rather it reflects a broader and deeper capability for comprehending our surroundings – ‘catching on’, ‘making sense’ of things, or‘figuring out’ what to do” (Gottfredson, 1997a).
As I hope you'll have noticed by now, the common theme throughout these definitions is UNDERSTANDING, and COMPREHENSION, if you have trained an AI to do a task based upon training alone, which is what machine learning is, then it DOESN'T UNDERSTAND independent of his HUMAN-DERIVED TRAINING DATA.
So no, GPT-3, nor any AI, that doesn't understand what it is doing, or understand the concepts it is using, is not intelligent.
Next time, just pipe down, you don't know shit.

[Anonymous No.12018431]

>>11992766
>>11993686
>>12007239
>>12015104
Looks like I need to whoop some pseudo-intellectual ass:
What is the academic definition of intelligence?
From the American Psychological Association (APA) Task Force on Intelligence:
>“Individuals differ from one another in their ability to understand complex ideas, to adapt effectively to the environment, to learn from experience, to engage in variousforms of reasoning, to overcome obstacles by taking thought” (Neisser et al., 1996).
Here’s a widely accepted definition among researchers:
>[Intelligence is] “a very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas,learn quickly and learn from experience … It is not merely book learning, a narrow academic skill, or test-taking smarts. Rather it reflects a broader and deeper capability for comprehending our surroundings – ‘catching on’, ‘making sense’ of things, or‘figuring out’ what to do” (Gottfredson, 1997a).
As I hope you'll have noticed by now, the common theme throughout these definitions is UNDERSTANDING, and COMPREHENSION, if you have trained an AI to do a task based upon training alone, which is what machine learning is, then it DOESN'T UNDERSTAND independent of its HUMAN-DERIVED TRAINING DATA.
So no, GPT-3, nor any AI, that doesn't understand what it is doing, or understand the concepts it is using, is not intelligent.
Next time, just pipe down, you don't know shit.

[Anonymous No.12018487]

>>12018431
>it DOESN'T UNDERSTAND independent of its HUMAN-DERIVED TRAINING DATA.
No human alive understands English, since we learned it from the training data provided by our human parents.

>doesn't understand what it is doing, or understand the concepts it is using
If it can write an essay indistinguishable from a high schooler's on a given topic, what evidence are you using to conclude that the high schooler's internal mental state while writing contains "understanding" while GPT-3's does not?

[Anonymous No.12018505]

>>12018487
>No human alive understands English, since we learned it from the training data provided by our human parents.
Wrong, humans intuitively understand concepts, AI doesn't. That's the difference.
Also, if what you said were true, we would have never derived the English language, as that requires understanding concepts from a neutral state of "no understanding".
Consider this a reply to your second line, as it doesn't really require addressing as you misunderstand the fundamentals of understanding.

[Anonymous No.12018510]

>>12018487
>>12018505
Also, this shows you didn't read the academic definitions of intelligence, as they make very clear their meaning of understanding.

[Anonymous No.12018573]

>>12018505
>humans intuitively understand concepts, AI doesn't
I'm asking you to defend this assertion with either evidence or some kind of logical backing and you aren't.
>if what you said were true
Obviously it's false, the point is that the AI's relationship to English is the same as most human speakers, who learn it from examples and don't derive it from scratch. Are you denying the intelligence of all these people? How does their use of language demonstrate an "ability to understand complex ideas, adapt effectively to the environment, learn from experience..." while the AI's does not?

>>12018510
If you're referring to the one that includes this line
>It is not merely book learning, a narrow academic skill, or test-taking smarts.
it's unfalsifiable as this states outright that it can't be tested. No matter what the AI does, you can claim it doesn't "reflect a broader or deeper capability for comprehending" because this is just vague bullshit with no objective criteria to distinguish "intelligence" from whatever it is you think GPT-3 is doing.

[Anonymous No.12018767]

While we're on the topic, my uni uses pic related for teaching the AI course. Is it any good?

[Anonymous No.12019644]

a

[Anonymous No.12020503]

>progress in AI will just suddenly stop right in time for it to not interfere with my life in any way
some of you are actually Dinosaur-tier naive.

[Anonymous No.12021832]

a

[Anonymous No.12021872]

>>11992176
>6.2.1 Gender, 6.2.2 Race https://arxiv.org/abs/2005.14165
Topkek
The AI is on our side, Anons

[Anonymous No.12022985]

a

[Anonymous No.12023193]

>>12007147
Your autism is set to 12.

Please dial your autism back down to a number on the scale of 1-10, sir.

[Anonymous No.12023292]

there needs to be a Universal Basic Income some time in the future and I think there will be

[Anonymous No.12023342]

>>11992176
I just want my robowaifu

[Anonymous No.12023366]

Why is this thread still here?
It really belongs on /lit/ not /sci/ since the way you trolls treat so-called inappropriately-termed 'AI', it's basically a work of fiction.
It's not much better fundamentally than what they had 20, 30 years ago, we just have bigger faster hardware to run it on, so they make it even more bloated than it already was.
Meanwhile it's output really isn't impressive at all.
Really, the current approach is a dead end, it'll never be real, general AI, not in our lifetimes.
We don't have a clue about actual 'intelligence' in the first place. It's all throwing things at the wall to see what sticks, experiments, guessing about it really.

[Anonymous No.12023641]

>>12023366
I think mostly everyone with a working brain knows AI isn't real. And anyone with knowledge of chemistry knows it is impossible in the future.
However with enough statistical learning and firepower the algorithms can solve/retrieve data faster than humans can. They may be able to do everything humans can actually since it's really just statistical. See all: https://en.wikipedia.org/wiki/The_Wisdom_of_Crowds

[Anonymous No.12023814]

>>11993714
has anyone in the history of man talked more out of their ass than you?
>communsim is direct result of protesting automation

what the fuck

[Anonymous No.12023831]

>>12023342
this desu iktfb

[Anonymous No.12023859]

>>12023641
I just don't agree. There are things that a small child knows the difference between because there is actual cognition happening, meanwhile the snazziest so-called 'AI' has no cognitive ability whatsoever and wil always get it wrong. If 'training data' is all the software has to go on then it will fuck up something that any human would get right. That's why it's garbage.

[Anonymous No.12023862]

>>11993714
>>12023814
Communism is one of those concepts that looks nice on paper but when you add the human element to it, with all it's flaws, it becomes a shitshow real quick, which is why China is such a shithole.

[Anonymous No.12024229]

MY HOBBY: Intentionally poisoning AI training data so they fail miserably at even simple tasks, revealing their fatal flaws

[Anonymous No.12025161]

>>12023342
Time to make one yourself.

[Anonymous No.12025272]

>>12024229
Based.
>>12023292
100% this and still absolutely hate the modern communist/marxist movements
>>12023862
>when you add the human element to it, with all it's flaws, it becomes a shitshow
Which is the point here. There is no "human element" when the means of production become fully automated and nobody has to work if they don't want to--because nobody's work would be economically valuable anymore in the first place. The entire tradition of criticizing communism is really as you implied just that it requires post-scarcity to work. But once you have actually achieved post-scarcity, then there's no sane alternative to absolute socialism or "Communism"
some cheeky people came up with the term
>Fully Automated Luxury Communism
but we can think of a better one later...

[Anonymous No.12026393]

>>12025161
AHHHHHHHHHHHHHH

[Anonymous No.12026472]

>yeah.. let's just run this regression over and over..

VC's (and reddit): ZOMG IT'S AI LIKE TERMINATOR AAAAAAAAAAAAAAAAAAAA

[Anonymous No.12026496]

>>12020503
>t. brainlet who has never heard of the first two AI winters.

[bruce787 No.12026501]

>>12024229
all types of data is valuable for the AI to ingest.

t. career in automation

[Anonymous No.12026529]

>>12025161

I HATE being offered food that I don't want and then guilted into taking it and then throwing it away secretly if there's any possible way I can avoid eating it. So I hate this robot.

[Anonymous No.12027623]

>>12023859

Well look at the deep fakes for example. Making fake anything always took cognitive ingenuity.
An embarassing picture of a nobody would be said by that nobody to be photoshopped. Though not taken seriously because no one would spend that much mental resources to fuck with an individual. Now through sheer statistical methods and enough algorithms it can be done.
So AI doesn't exist and never will but it is a threat.

[Anonymous No.12029148]

>>12026529
If it's something small, just eat it. Wasting food is worst than not eating a small portion being offered.

[Anonymous No.12029946]

>of the inevitable upcoming AI revolution

Sounds like the milionth thing predicted by muh science that will never happen.

[Anonymous No.12030444]

>>11992756
even replika shiws more sentience than most humans online

[Anonymous No.12030482]

This is something that genuinely frightens me. It will all happen very quickly and without us suspecting a thing.