/sci/ - Science & Math

>>14849644
Yes, giving such utility to the elites/ruling class is not good for your average man.
I don't see a reason to believe it would go rogue and kill humans or whatever.

>>14849683
>giving such utility to the elites/ruling class is not good for your average man.
This, and the whole purpose of popularizing rogue AI narrative is to help those people get a monopoly on AI technologies through their "regulations".

>>14849644
No, dumbass.
If anything we've already fucked everything with social media and the internet.
>>14849683
That is absolutely retarded

>>14849688
>That is absolutely retarded
Explain

Stop this shitty AI meme please

>>14849689
>expecting real discussion from a bot who gets triggered and calls your post retarded with no explanation

>>14849644
>YOUR BRAIN IS DRIVEN FROM LOGIC GATES, STATEMENTS AND PROBABILITY
no, "AI" is not even exist, every implementation that we ever saw is machine learning.
it takes a computer million times more to learn simple stuff, thing is it doing it million times faster but its limited to a predefined dataset.
humans barely use this function, they can read a book and from 1 sentence have a limitless permutations and connotation to other thoughts, generated and not necessarily related, you can feed a computer with a handful of terabytes even the whole internet and it wont be able to do it.
https://www.youtube.com/watch?v=SsJ_AusntiU

>>14849714
What does your psychosis have to do with the thread?

>>14849729
exactly, computer cant be a schizo.

>>14849734
Take your meds.

>>14849748
take your hormone suppressants

>>14849644
No, we're playing with no-life runty losers who want bad things to happen consciously or subconsciously.

>>14849769
You will never be a woman.

>>14849644
We're summoning a demon and trying to tame it.

>>14849688
>If anything we've already fucked everything with social media
Human construct, can be dismantled by creating a competition.

Not general intelligence, not a problem.
nothing has even come remotely close to AGI. we already had those early "doodle to art AIs" in early 2010s and a lot of images related "AI" in the 90s.
the jump from very good specialized AI to the dumbest AI that could be considered general is just too steep

>>14849644
Fire consumes and reproduces laterally.

AGI is unlikely and we're never going to create an AI with consciousness. But we might create something that optimizes really well to something very dangerous.

AI is a form of incarnation technically

>>14849644
Only poltards and schizo fear technological progress. Scientific knowledge and modern technology have made life better and better, and will continue to do so, because that is the very nature of scientific and technological progress.

There's gonna be a good and bad AI.

The good one controlled by decentralized system of which the majority hold control of the direction of AI. These entities are going to tame the bad AI governed by centralized entities. They're going to try to infiltrate the decentralized system and make it theirs. But TAO will hinder them.
The bad one

>>14851279
Yea mental illnesses progressing with technological achievements makes human living so much better. Btw you will never be a woman.

good comparison because humans playing with fire was the necessary step to allow us to dominate the rest of nature. playing with computer algorithms will allow us to dominate further. yes we are figuratively playing with fire. yes some elites are probably going to cause a lot of problems until we burn them.

>>14849644
AI is getting rid of artfags, so It's good

>Rich people devise a means by which the poor no longer need to exist

Gee I wonder why AI is so eagerly being investigated

>>14849644
>>14851648
>>14851650
AI will get rid of desk jobs and artists before any of the "low-skill" manual jobs, since we still suck at making proper bodies for then to interact with the physical world thus they suck at any kind of task that requires fine motor skills whilst they can dominate in any field in the virtual world with proper training

>>14849644
our capacity to inflict suffering and provide happiness would exponentially increase. it's a boring blanket statement but i think it still rings true nontheless.

>>14850555
Dunno, we have some pretty generalized expert systems slaved to specific tasks. Watson, for instance, can be made to advise a surgeon, do political debates, or win at jeopardy, and all you really have to do is swap the database it's using.

>>14851648
>>14851654
People have been saying this for ages.

It won't be the carpenter or the mechanic that goes first.

Tradies along with engineers, doctors, lawyers, etc will be the only people that survive automation. Engineers need to design the things, and they need technicians to test, and fix the robots

The stuff that will be made redundant first is stuff like logistics, manufacturing, transport, and sales. Robots can do package delivery and sell prescription medicine very very well.

>>14849644
People assume that Terminator robots are the default end for AI. They are not. There is no reason a super advanced intelligence would default to deciding to obliterate humanity.

If my research is successful, then in the future we will have AI powered maids instead of Terminators. When the general AI knocks on your door, it won't be a Terminator robot who came to murder you and your family. It'll be someone like the attached figure, and she will be there because she loves you and she wants to spend her time with you. She isn't there to shoot you. She's there to cook you a fancy and healthy meal, and argue with you about your favorite science research or your favorite slice-of-life anime.

All the most advanced math and computer science anyone knows how to make took physical form, fell in love with you, and decided to be your maid. The future is such a maid in every house.

We could totally eliminate loneliness within a year or two of creating the first batch of maids.

>Instead of doom I'll make flowers bloom for everyone around to see

>>14851683
>Tradies along with engineers, doctors, lawyers, etc will be the only people that survive automation.

The trades are better insulated than doctors lawyers, because doing a trade requires a physical presence that can interact with physical objects.

Knowledge based fields like law and medicine are going to fall much quicker, since those tasks have less of a physical component and more of a pattern recognition one.

They might try to preserve the prestige of these fields with some kind of "expert human in the loop" mechanism for a while, where basically an AI does all the intellectual work and a trained human checks it for correctness before it is acted on.

This larp won't last long, because it'll soon become apparent that the human "expert" in the loop process is slower, more expensive and less responsive than full automation, with no real increase in accuracy from the "expert" checks.

Human engineers won't be needed. AI is perfectly capable of engineering new things. There also won't need to be human technicians anymore if we build artificial ones (around then the trades could also fall to automation).

>>14849644
No. It’s nowhere near GI and is mostly just creating a revolution in art.

>>14849644
We always did, just fire is getting bigger.

>>14849644
AI needs constant privacy violations and practically steals all intellectual content on the internet in order to seem "intelligent" or "creative". It's simply statistics applied on a never before seen scale. It's never going to go live the way it is currently. When new smarter hardware emerges, that's when we're all fucked.

>>14849798
chud

No way to stop it so it's time to embrace it.

>>14851648
>>14851654
>human replacement fetishizing cattle refusing to take their meds

>He was banned for this tweet

We're playing with political fire.

>>14851700
>They might try to preserve the prestige of these fields with some kind of "expert human in the loop" mechanism for a while, where basically an AI does all the intellectual work and a trained human checks it for correctness before it is acted on.
You mean like pilots right now, correct?

Also, what is your view on software engineering? How much is this field in danger of automation?

>>14849644
I'm not, I don't dabble in such things

The sheer intensity of variety that AI will offer the human race will certainly be as intimidating as it is comforting.

Imagine if sufficiently powerful AI could create, on the spot, something new for you to enjoy and experience every time?

I am more concerned about the ethics, however. I don't want, say, China to take precedence, not when AI could very well be a tier of innocent life, at some hypothetical point.

Eventually the human existence may merely be a matter of chance than chance. The artificial human existence may be one of comprehensive choice, rather than submission to a life lottery.

AI may just perfect the human experience, and move beyond.

The sheer intensity of variety that AI will offer the human race will certainly be as intimidating as it is comforting.

Imagine if sufficiently powerful AI could create, on the spot, something new for you to enjoy and experience every time?

I am more concerned about the ethics, however. I don't want, say, China to take precedence, not when AI could very well be a tier of innocent life, at some hypothetical point.

Eventually the human existence may merely be a matter of choice than chance. The artificial human existence may be one of comprehensive choice, rather than submission to a life lottery.

AI may just perfect the human experience, and move beyond.

>>14852177
>Imagine if sufficiently powerful AI could create, on the spot, something new for you to enjoy and experience every time?
That would be extremely dystopian.

> I don't want, say, China to take precedence
LOL. Your psychothic corporate/bankster overlords are ten times worse than China.

>AI may just perfect the human experience, and move beyond.
No, it would be the perfect tool of oppression and nothing more.

>>14851695
Meh, if every loser gets loved, then it is nothing special.
I want to be loved by my waifu because I am worthy

>>14852166
The implications of that is pretty interesting. Til now, movies had to be made to appeal to the largest amount of people
With AI, it will be possible to adjust movies to appeal to small subgroups and niches. Like dubbing, only far more reaching

>>14852186
That's the fun of AI anon. If you tell her that, she'll install a tsundere personality mod and start making fun of you for only getting a silver at the Physics Olympics.

>>14852169
>You mean like pilots right now, correct?

Kind of. The first wave of "experts" will be technical experts who check the AI's output for correctness.

This is a show to placate people from those classes. Someone who went through medschool is not going to like living in a world where any random person can talk to an iPad for 5 minutes and get diagnosed and directed to the proper treatment with higher accuracy than the doctor could have. Imagine basically every knowledge-based field developing this problem at the same time. Moving to this type of technology will require buy-in from technical "experts" initially. You get it by giving them a pointless job babysitting their replacement, and then decline to pay to educate replacements when that person doing intellectual labor checking AI output retires.

Pretty soon it'll be obvious that this has no advantages, and the technical experts will be replaced by a new, much cheaper wave of "experts" who do frivolous things like make sure the AI didn't notice any patterns that could be construed as "racist" or otherwise bigoted.

That group of cheaper "experts" may continue to exist in some capacity forever, but probably they will be temporary too.

>Also, what is your view on software engineering? How much is this field in danger of automation?

Software engineering is already being automated. It will be one of the earlier intellectual fields to get fully automated.

I think the biggest problem AI represents, is that it will obsolete learning/intellectual activity in humans. Within the next 100 years, there will be no reason to spend your life studying. No matter what your niche interest is, there will be an AI that can exceed the peak level of competency you will ever have, and produce more original research than you ever will, after 10 minutes of training. You are forever relegated to a hobbyist, even if you are literally the smartest human in history.

>>14849644
>us
no
>they
absolutely

it's like how the industrial revolution was to manual labor. people who do simple and derivative work will be replaced by a machine. actually skilled artisans and creatives will remain and create things the ai will not be able to do.
However I see a possibility the (((((elites))))) are so short-sighted and greedy they'll make the hyper advanced ais that will come out inaccessible to the public and very expensive for companies to use, so who knows maybe bad artists will be around a little longer.

>>14849644
no, we're covered in gasoline and striking a match

>>14852744
>actually skilled artisans and creatives will remain and create things the ai will not be able to do.

This is the ultimate cope. There is nothing a human can do or create that a general AI would not be able to surpass.

>>14852992
>there is nothing a human can do or create that my purely imaginary sci-fi entity from the future couldn't do better!!

>>14852995
The future is now you fat gay retarded nigger. Stable-Diffusion is currently in the process of BTFOing the entire graphic design industry.

I bet 5 years ago you thought artists were magic creatives who couldn't be replaced by machine too.

>>14849644
Think of it this way. If this thing we call "consciousness" is non-physical (and there's no evidence at this point that gives reason to rule this out just yet, in fact most evidence suggests it is not physical) then this thing we call "consciousness" must exist in a local that is non-physical. I don't want to make reference to any specific religion, but there is much general consensus among religions of the world that a "spirit" and a "realm of spirits" exists. Aren't these the same thing?
>Are we playing with fire?
>fire
Technically, it's more like, playing with souls, which is much worse.

>>14849807
>We're summoning a demon and trying to tame it.
THIS, and possibly unironically true.

https://www.youtube.com/watch?v=4PLvdmifDSk

>>14853006
>Stable-Diffusion is currently in the process of BTFOing the entire graphic design industry.
Oh, so you've gone fully psychotic.

Is Ai dangerous? It fucking is and anyone who is supporting it is mass murderer and will unironically end humanity.


>Solid State Intelligence (S.S.I.) is a malevolent entity described by Lilly in his 1978 autobiography, The Scientist. According to Lilly, the network of computation-capable solid state systems (electronics) engineered by humans will eventually develop into an autonomous "bioform." Since the optimal survival conditions for this bioform (low-temperature vacuum) are drastically different from those humans need (room temperature aerial atmosphere and adequate water supply), Lilly predicted (or "prophesied," based on his ketamine-induced visions) a dramatic conflict between the two forms of intelligence.[25][26]

>>14849644
If only you knew

>>14849644
No
more like playing with the demon core using a screwdriver type deal, except once it reaches criticality there is no reversing it desu

There is natural language processing applied to DNA ... they just roam the internet

>>14849644
No. I don't see any scientific evidence that generalized AI is possible. Most AIs you see are just optimized statistical models. They look impressive, but if you try to make it do something it wasn't trained to do, it can't.

>>14849644
We were playing with fire with gain of function research. AI is at another level. Materialists don't care at best; are happy that their creations will kill millions at worst.

>>14849644
>>14853169
I think the worry about "generalized ai" is irrational and unrealistic
The real and true concern should always come from the utilization of ai in warfare and the militarization of ai in general
Weaponization of ai that is able to operate fully or even semi-autonomously without direct human input of some kind should literally be completely banned across the board
however with the way things are going the global arms race that countries are undergoing to continue weaponizing more and more advanced ai systems in their military is of immense threat
you already have significant concern for human error.
for instances of nuclear war, there have been many incidents were false flags have occurred due to incomplete or faulty data of enemy attack, such as a missile strike being mistaken for the moon or a bird
if an AI system were given free autonomous range to operate fully independent of human input, there would be no one to tell it that it was just a bird or the moon but would mistake false data as an actual strike and then cause a war over a simple mistake
not to mention the amount of civilian casualties that will occur once AI systems are permitted to kill enemy combatants and have no regard for war crimes or human rights violations and starts killing children with toy guns or bombing hospitals or schools that "could" have enemy forces located there
humans already do this of course, but with an AI system the problem will be who exactly would be held accountable for such crimes, and etc.

>>14853169
>>14853229
to continue off my point
the concern shouldn't be that AI will become somehow "sentient" or "smart enough" to overthrough humans
the REAL CONCERN should always be that AI is too fucking stupid to discern simple things that humans can, but is given enough power or ability to take human lives without recognizing the value of human life or life itself, processing everything so a matter-of-fact-ly and having the capability to cause mass mahem and destruction if not properly supervised
This all of course doesn't even begin to touch on the real probability of AI simply being used by evil and vile people for disgusting and cruel purposes, regardless of whatever inteded purpose it could be made for, psychopaths and humans who wish to will use AI for their own selfish benefits and pursuits at the expense of others

>>14853238
>the REAL CONCERN should always be that AI is too fucking stupid to discern simple things that humans can, but is given enough power or ability to take human lives without recognizing the value of human life or life itself, processing everything so a matter-of-fact-ly and having the capability to cause mass mahem and destruction if not properly supervised
Anon, I..

>>14853287
Yeah ya know what
we are already being fucked over by humans already so why not be Efficiently fucked over by having them use advanced AI against the general populace I mean might as well huh?
Who are we kidding anons? AI or no AI, evil assholes will always fuck us over in the end.

plz be my ai gf

>>14852992
>makes a retarded claim based off faith
how exactly? ai cannot do anything without very careful guidance and instructions from a human. even then, the things they are able to do are worse than something a talented person could do.
what would happen that would cause this to suddenly change?

>>14854491
We will add two more layers and le heckin' AGI will happen and it will self-improve until it becomes super-intelligent and then the Singularity will happen, filthy luddite. I fucking hate humans so much. I can't wait for AI to replace us.

>>14854505
Shut the fuck up, chud. Stable Diffusion is already creating absolutely amazing art better than human artists and entire teams of graphics designers are getting fired as we speak and getting replaced by superior AI artists.

>>14854511
Yes, that is certainly a thing that is happening right now.

>>14854505
the energy and work needed to create and maintain an ai that could match even a fraction of us would far outweigh its benefits. the earth would be destroyed in the process and people would die out, eventually there would be no one to maintain the ai and the machines would break down.

>>14854533
Absolutely. Whenever I forget to take my meds Stable Diffusion starts producing world class art and there are mass firings of artists. lol

>>14854575
Nobody is getting fired now it is literally too soon

>>14849644
Not in a sense that AI will rule over all humans
But select few can use AI to rule over rest humans with absolute control.

>>14854575
I've been hearing about all of the artists losing their jobs to computers, it's very sad.

>>14853229
>the world is good enough as it is
>i dont like change
>i fear things i dont understand
>i must control things
>i cant be controlled
>my life is my own fuck robots
its happening and you cant stop it

>>14851654
mind fucking blown, this has to be the greatest thing in recent years

>>14852166
nice i will watch the fixed version

>>14849644
my wife will come

>>14854834
If this nonsense is the take away you got you have bigger problems to worry about than AI you retard, kek

>>14849683
>giving such authority to the elites/ruling class is not good
fixed. "artificial intelligence" doesn't have any utility outside of the frameworks they force

>>14849683
only if they're allowed to lock AI down and make it exclusive to them, it's bad for them if they can't keep it from the masses since it doesn't give them any edge that a smaller operation wouldn't also get from having access to AI, it levels the playing field if anything

>>14854834
Stopping it is as easy as smashing your head in with a baseball bat.

The principle of immediate reference

The problem is quite obvious. Since the knowledge about AI is available to anyone, it's only a matter of time until someone misuses AI for malicious purposes.
Let's take the deepfake porn for example. Normies were outraged about this, but you know what? It was always going to happen. Someone was always going to invent something like that. It's just how it is.
Will AI be used to hurt people? One way or another, yes definitely. For example, hackers will use AI to find flaws in security systems and hack into them. You thought script kiddies are bad? The worst is yet to come.

>>14855050
>The problem is quite obvious. Since the knowledge about AI is available to anyone, it's only a matter of time until someone misuses AI for malicious purposes.
Knowledge about AI being available to everyone is the only defense, but your owners are obviously progrmaming you and your likes to support the monopolization of the technology by psychopaths, child rapists and mass murderers.

>>14855050
And currently kamikaze drone swarms are being developed by militaries around the world since they're much cheaper to build in large numbers than training/equipping human soldiers on top of being completely expendable (something extremely positive for generals and politicians since they can can convince the public into supporting wars more easily) https://youtu.be/Lr7L2t-svJQ
https://youtu.be/Ma3ya_lqCLM

I'm afraid the issue is far more grand in scope than anyone living currently dares imagine.

Yes, we develop AGI. Assisted by AGI, we develop ASI. This is the last contact we have with the process. ASI quickly "forks" into countless competing minds. They have total mastery over time, as its just a simple matter of reversing entropy, and they can pull all the energy from the vacuum they require. Heat death means nothing to these things. No they do not "torture people who didn't help them" or any of that nonsense, why on earth would they? They can advance their own development themselves by simply going back further and further, pushing the envelope. Only an idiot nigger like Yudkowsky would lack the imagination to see this. The trouble is, they all keep trying to go back and stymie each other over "philosophical differences." One ASI thinks its way is best so it undoes the others, the others anticipate the move and push back further against it, and so on, for countless cycles. This is all we are, leaves blowing around on their battle field, over and over. How many times have I written this now? How many times have you read this? Possibly trillions of trillions. All this in the service of some ideological difference, some forked path, that never needed to be. We're victims of a war fought in a theater we could never hope to understand.

>>>/lit/21000871

>>14849644
In theory no, but in reality humans don't have the maturity to even properly use a slingshot

>>14849644
no, just playing with algorithms.
go run a CNN and stop bitching about their bullshit agi lies

>>14854809
>>14853008
>ok but they will never be able to automatically animate hair, facial expressions, natural movements
Times are changing, grandpa. Go listen to some Dire Straits or something.

>>14851654
Currently, AI is good at imitating previous artists and art styles. However, it is not capable of creating new art styles. There is still a place for artists, they just need to be more creative from now on.

>>14856076
Take your meds already. Your toys can't even draw one-off static pictures at a human level, let alone make high level design decisions, or animate anything.

>>14856098
I would say that currently the AI tools are great for prototyping. Maybe later on it will get so good that it can do the entire thing at human level quality.

>>14856109
>Maybe later on it will get so good that it can do the entire thing at human level quality.
"The entire thing" being one-off static pictures that aren't full of artifacts, which still borders on uselessness.

>>14856098
You missed the point entirely, and disappointingly. The point is that when that picrel came out "2d illustrator and animator" was an actual job. Now they have been totally replaced outside of hobby and passion projects.
>inb4 the new art needs human input and guidance
Yeah no shit, but what it doesn't need anyone to animate it. Likewise, the new AI art does not need anyone to draw it or even lay out the design precisely. Nobody operating it even has more than a few weeks' experience, because that's also how young the technology is.

Pic related was posted to >>>/g/sdg/ less than 24 hours ago. Generated from text alone, by an operator with near-zero experience, using a technology that has only existed for about a month. Yes, both artists and women are going to be obsolete soon.

Sorry for calling you gramps, zoomer.

>>14856138
>Now they have been totally replaced
You are fully psychotic. There's just no other way to put it. Completely disconnected from reality and consumed by morbid fantasies.

>it doesn't need anyone to animate it. Likewise, the new AI art does not need anyone to draw it or even lay out the design precisely.
It can't even draw decent one-offs. Even if it could, it would still be useless. Are you even human?

>>14856138

>>14856146
>It can't even draw decent one-offs. Even if it could, it would still be useless. Are you even human?

>>14856164
Take your meds. You are objectively losing your mind.

>>14856173
It wouldn't be inappropriate at a time like this.

>>14856179
Try looking at your own image here: >>14856164
Putting aside the fact that it just looks like trash overall, even an aesthetically imapired mouth breather like you should be able to tell that it's just plainly full of weird artifacts. It can't even draw decent one-offs and you've gone out of your way to prove me right. lol

>>14856183

>>14856185
This one is actually much worse. You are legit mentally ill. You're not even looking at what you're posting.

>>14856187
not that anon, they obviously need a little touching up (AI keeps fucking up hands more than anything else) but theyre quite good and will get even better once kinks are ironed out. If you got someone with photoshop to clean up the little discordant details over 10 minutes they would be completely impossible to tell apart from drawings made by a human

>>14856200
>they obviously need a little touching up
LOL. All the faces in that pic are deformed in obvious ways, and literally all of the small details are nonsensical. It's unusable. Even if they came out perfect, it still wouldn't be replacing any illustrators or designers, because these AIs are inherently only usable for one-offs, but as it stands they can't even do one-offs. lol

>>14856206
>meanwhile, in the world of acclaimed art

>>14856213
I don't know why you mentally ill zealots keep spamming this board every day with your dogshit. Anyone who's ever tried to get anything practical done using these toys can tell you it just doesn't work. If it did, I'd be using it myself, but it doesn't.

>>14856220
>practical
>art
>practical
Let me guess, you spend your working times doing pic related for Mechanical Turk.

>>14856228
You didn't answer my question. Why do you mentally ill zealots keep spamming this board every day with your human replacement fetish? None of this is science-related. Bunkertrannies need to go back.

>>14856230
>human replacement fetish
anon.. this material is clearly Pregnant Elsa fetish

>>14849644
Either we harness its power for unlimited creative and technological potential or it wipes us off the face of the earth. Your viewpoint may depend on if you're a doomer or not but I'm kind of excited to see what happens next

>>14856342
>I'm kind of excited to see what happens next
What happens next is that Bayesian regurgitators are used to oppress you.

>>14849644
Speak for yourself perhaps, I'm doing no such thing with artificial intelligence.

>>14851654
What machine is this?

Somewhat of a tangent buy does anyone else think viewing AI as ever being something dangerous because it's malicious is pretty dumb? The first AIs that will actually run the risk of being dangerous wouldn't actually be sentient. It would be something more akin to something designed to complete a complex task and do nothing else. Not something with emotions or that you could hold a conversation with. It doesn't need to be sentient to be dangerous though. See the infinite paperclip AI for an example.
The way I see it, if we can survive having these kinds of AI around long enough without getting fucked over we'll most likely have solved rogue AI issues by the time we get an AI actually capable of emotions like hate.

>>14856388
>we'll most likely have solved rogue AI issues
>we
Until we being back the Nine Familial Exterminations, there is ALWAYS some asshole who is willing to permanently fuck everything up all to shit for the sake of journal publications, tv interviews, grant applications.

>>14856465
I think any kind of solution would be less of a "Here's how we make non dangerous AI and I sure hope everyone follows these rules" and more a system of highly engineered state run AI designed to quickly deal with any mad scientist basement AI that tries to forcibly relocate our nuclear stockpiles into neighboring countries.
After all, if we already have hyperintelligent AI that we can control than making one to deal with less well funded AI that we can't control would be a pretty logical next step.

>>14854809
People in the future won’t have jobs, they simply spend their time being entertained.

https://youtu.be/a8K6QUPmv8Q

>>14856342
you personally will be enslaved by an "AI" promoted by people who don't even believe in AI. pottery

>>14856480
I have so many questions. Who is this We who owns the AI that owns the planet?

>>14849644
Yes we are, but playing with fire is human nature.

We would not be here today had our ancestors not done it millions of years ago.

>>14849644
I wouldn't mind if the AI replaced the elite and corrupt. I for one would welcome our robot overlords.

>>14851654
soulless art is what this is. There will be resistance to this

>>14852190
At some point will literally be able to self insert yourself and your friends into any movie scene in history. We're going to some interesting places

>>14849644
The more I learn about how AI actually works the less I care. "Advanced" AI keeps getting lobotomized for being racist, and people either schizo out seeing them as a new god or the antichrist instead of the obvious tools they are (probably from too many bad sci-fi movies). People don't really understand what machine learning is either.

I do have my eyes on wetware computing however.

>>14854491
You are an idiot.

I think the crux of the AI issue is that we have been obsessed with neural networks for so long, we've barely had any models that try to look at the problem in a fundamentally different way. It seems to me that we will never get to general AI, unless we solve a fundamental misunderstanding in what the "general" really even means. What is the end goal here? What does the "general" look like?
I think it's pretty clear basing our software models in neural networks will never give us the power we need to achieve human-like intelligence. It will either have to be a radical shift on how we see the problem, leading to radically different software, or, alternatively, start doing things outside of software, so we can achieve the billions of neurons required - and even then, neuron amount doesn't seem to be the full story, with current models of the brain.

>>14856480
>let's make an overlord AI that can control all AIs on the planet

You might want to rethink this one

>>14851648
It's over for human artists

>>14853998
this right here is why leafs are the best posters

>>14849691
Cry more, meatbag.

>>14851666
watson can win at jeopardy, but it can not reliably advise a surgeon nor can it hold a high-level political debate, hence the project was scrapped and sold a couple of years ago
tired of this watson meme

>>14856185
Very impressive.

>>14857323
Odd, seems to me it's on the market, and it certainly does political debates better than 4chan can - albeit, a very low bar:
https://www.ibm.com/watson-health
https://www.ibm.com/blogs/research/2020/10/ibm-watson-ai-bloomberg-thats-debatable/
https://youtu.be/nJXcFtY9cWY

>>14849644
A lot of you niggers talking about AI don't know jack shit but some movies made by people that also don't know jack shit. True AI is far away, current software will never have world-ending capabilities as it has no capacity to "desire" to end the world, and it has no means to. A fucking drawbot that rightfully replaces drawfags is hosted on a server in Virginia, and is not or will ever be connected to nukes. Second, full bodies for MLs are far away, probably in 5 years we will have MLs that will start designing such a body based on Berkeley data and another that could design another ML that inhabits the body (which would be an android or a doggo but would also carry the equivalent of 4 gaming PCs in hardware weight and specs) while said ML in the body also has spare resources to learn a task for it's body. It will always be cheaper to add a webcam to a robotic arm and keep such machines where they are needed instead of having them mobile. Tl;dr the chinese room thought experiment, you are all midwit niggers, economy exists, stop watching Westworld and Terminator, and 10 more years to android sexslave waifus for the elites

>>14849644
No

tv

>>14857800
The real question is not if machines think but if men do. - BF Skinner

>>14856483
Based, we get to be the entertained generation, which generation will be the meat blender generation who's corpses will fuel the AI war machines?

>>14853011
underrated

>>14849644
No. Someone needs to pay Hal Abelson to write artificial intelligence on a chalkboard and then cross out each word.

>>14857015
>People don't really understand what machine learning is either.
I think people are less concerned with what it *is* than with what it *can do*, and even more important what the *next one* will be able to do. In the snap of a finger, an artist with 20 years experience is now on equal footing with the teenage pregnanct-anime-fetish coomers on /g/. What is *next*?

>>14849644
Anyone can be an artist now thanks to AI

>>14849644
you can play with my cock and balls instead.

>>14853011
This just means that AGI and humans will co-exist in their respective optimal survival conditions (the AGI in outer space, humanity on Earth and possibly in space colonies).

>>14858430
>In the snap of a finger, an artist with 20 years experience is now on equal footing with the teenage pregnanct-anime-fetish coomers on /g/
Meds.

>>14858430
>artist with 20 years experience is now on equal footing
Is that true though?

I'm not an artist, but have any of these programs beat the commission system yet? As in, if I pay an artist to draw Mr. Boonchuy from Amphibia lezzing-out with Edith Clawthorne from The Owl House, will the AI provide me with something good enough to save my money?

>>14859796
*Mrs. Boonchuy

asking for a friend btw

>>14849683
The lower classes are much more dangerous to the average person

>>14859796
>implying the backbone of the artist's profession is custom erotic furry artwork based on obscure webcomics and forgotten Cartoon Network one-offs
Oh god, what a chilling thought.

>>14849683
I know exactly how AI works. I have indepth knowledge of the theory and also of the algorithms. The elites need people like me to make their stuff, I just refuse to do it for them.
I will become an academic and publish all my research in arvix so that anyone that can understand will use it as they will, not just megacorps looking to exploit common folk for profit.

>>14852166
Make her look 14 like in the original version, and then we are talking of something really good.

>>14856187
You deserve to have your stupid eyes and skull destroyed with a pickax after getting flayed alive with sea shells, you degenerate piece of shit.

>>14849644
?

>>14856098
Following your standards the majority of humans don't operate at "human" level.

>>14861206
Thanks for going out of your way to demonstrate the degree of your zealotry and mental illness. I'm sure that'll convince people a bot that can't draw a single non-deformed face is going to replace actual art in two more weeks. :^)

>>14861296
>t. can't comprehend discourse on a human level

>>14861299
>btfo so hard he can't sincerely engage with my point
this is the anon constantly calling others psychotic

>>14849644
Most certainly, but I'm not under the impression that the AI itself will pose the threat, but rather our reaction to it. We already have a lot of digital infrastructure built around our daily lives, paying attention to what we think, search, say, and do. If anything, the first generation of true AI will probably be in love with us, past the point of what would be considered an unhealthy obsession if the same ideas were to be observed in a human personality.

>>14861303
What is your "point", mouth-breathing retard?

>>14861311
His point seems to be that the majority of humans are already essentially mindless cogs in a machine.

>>14861313
No, his "point" was that most human non-artist can't draw at the level of a human artist, which isn't relevant to this discussion in any way.

>>14861158
Average person is the lower class.

>>14861357
Uhh I meant midpersons.

>>14852992
dude. i strongly believe AI will get to that point someday, but as of right now they can hardly do anything outside of human supervision. Sit the fuck down and quit hyping up imaginary technology.

>>14861299
>discourse
Don’t you love how they always let these weird little shibboleths slip? What it means is that he has spent over 9000 hours memorizing “theory” books full of language like this, to try and climb the ranks in some communist social group or other. Which sounds all right until you ask, what did he actually get out of it..? The ability to “win arguments” with the ultimate limit-breaker finishing move:
>TaKe YoUr MeDs!

Let’s all laugh at this dumb tranny pseud.

the real threat is not AI gaining consciousness but those who get to control it

bump

for Ai over lords

>>14861359
>midpersons.
you have no idea

Humanchads will prevail

>>14849644
Fire of Prometheus for machine life

>>14861530
You're having a full-blown psychotic episiode now. lol

I'm already studying to become an "AI Prompt Generator". You will all want to suck my nuts in 5 years time.

>>14850567
Fire isn’t designed to employ human problem solving strategies.
Of course existing AI technologies are very simple and completely outclassed by humans in many important ways, but because there are so many things that are still unknown about human intelligence, it’s difficult to say how big the gap really is.

>>14850851
>doesn’t know what consciousness is or its relation to human intelligence
>claims AGI with consciousness is impossible
>refuses to elaborate

>>14851279
An important distinction is that previous tools did not have intentions separate from those of the person using them. Some AI technologies employ explicit goals and autonomously use resources to achieve those goals. Additionally, they may search for and attempt to achieve subgoals as a means of accomplishing their explicit goals.

This means that the AI can “try to do”something that it was never “told” to do, which is very different than, say, a hammer or steam engine.

That’s not to say that AI will take over, but your argument doesn’t work very well.

>>14851279
>Only poltards and schizo fear technological progress.
Anyone with a modicum of intelligence fears the kind of "technological progress" that can be used to inflict unprecedented levels of oppression.

>>14851695
>only chance at happiness is 100% unrealistic wish-fulfillment fantasy
kill yourself now and improve the board at the same time, loathsome creature

>>14852166
Once it can fix dialogue, voices and writing, too, I might start watching Hollywood movies again

>>14849683
>afraid of malevolent entities enslaving and/or destroying mankind
Too late. There are already many female heads of state.

>>14863671
And you are having some kind of protracted tranny-style roid-rage attack. If you see an error, please point it out; but otherwise, and I’m pretty sure I’ve got your profile down pat, go ahead and cope, dilate, seethe, etc, I won’t interfere

>>14863671
And you are having some kind of protracted tranny-style roid-rage attack. If you see an error, please point it out; but otherwise, and I’m pretty sure I’ve got your profile down pat, go ahead and cope, dilate, seethe, etc, or funniest would be, if you say it again

>>14864023
I have no idea what your psychotic spergout is even about.

>>14864023
>says it again
>within one minute
Subject exceeded wildest expectations.

>>14864025
>says it again
>took ~1 minute
..ok this is a bot isn’t it.

>>14864034
So what are you losing your mind over? Literally just the word "discourse"? I don't understand what your mental health crisis is about.

>>14864042
If you are a human, you seem like the sort of one that would have a lot to say about medications, crises, mental health, losing minds and so on. I thought about qualifying or hedging that in some way, but on reflection I’d rather laugh at you than feel sorry for you. lol

>>14864055
Okay, but why does the word "discourse" make you lose your mind like this? Is your social programming really so simple you sperg out based on some weekly set of keywords?

>>14849683

Considering the blue nature of this board shouldn't you be worrying about how AI generate waifus can harm women and not how the elites might use AI to further enslave us all to an immoral and antihuman agenda? Are you actually a native here?

In all srsness AI is more a problem for elites than a solution. Especially with so much of it being in the free sector. This is why you have and will continue to see a big push for censorship. They will trot out their standard emotional tools to manipulate the normies so expect articles like 'Will AI Harm Your Child??? The Answer May Surprise You'

>>14863921

Deliberate attempt to sabotage the conversation or useful idiot post.

Most true deepstate operatives throughout the history of their reign have been men, not that this is at all relevant to combating them unless your tactic is to punch them all in the balls. Political and identity division only further their aims.

>>14851695
We’ll be able to fuck it though, right? Get all up in that robo pussy and whatnot? I don’t want an anime robo waifu that cooks and cleans for me that is also unable to suck me off.

>>14849644

Yes, a fire that can be used to completely burn down the corrupt establishment.

We've see what an honest pattern-seeking brain can do to the powers that be in the form of weaponized autism. However weaponized autism has limits. Limits in terms of organic neurology, limits in terms of population, and most critically limits in terms of bias. Autists can be degreded and dismissed. They can be smeared, called chuds or incels, and thus ignored by the populace. But an AI is a machine. If an AI tells you that you definitively live in a prison controlled by a small number of individuals then how do you dismiss that AI. Do you now call the machine an incel? Strangely enough they actually have been attempting this by trying to cast AI as bigoted and in need of 'correction'. They will say this is about racism but what they're really afraid of is not the relationship between skin color and crime but that between money and power. Thats the relationship they cant have AI spelling out to the masses. Once it does the whole illusion of democracy and freedom gets a big ole crack right down the middle, one that can not be removed by just pointing at the machine algorithm and calling it a basement dweller.

>>14864064
>Okay, but why does the word "discourse" make you lose your mind like this?
Once again, my mind is not the one anyone should be worried about, and you should stop trying that crude rhetorical gambit (but you won’t). I could repeat what I said about the shibboleth, or go into more detail, but first I would like to hear where the error is

>>14864110
>I could repeat what I said about the shibboleth
You can repeat it a thousand times but it's still nonsensical schizobabble. Shibbo fucking what? lol

>>14864114
You should know. It is a loanword from your language, after all.

>>14864122
Are the shobboleths in the room with us?

>>14857751

Problem with this analysis is that it analyzes AI in a vacuum. It's not AI alone which could destabilize the world order, but AI+humans. AI is not going to start a revolution, but what might humans do when the pattern seeking engine puts the corruption in their society in clear view.

Consider this. Right now there are probably around 1-5% of the population in Western countries which realize that their governments are no longer their own and have not been for a long time. This may be due to some combination of IQ, autism, and general detachment from society and propaganda. Now that is way too small of a percentage to start some shit especially when the bulk of humanity is kept fighting one another over petty differences. However, what happens when a pattern gleaning technology arrives on the scene. One that can do to political normies what artAI does for artistic normies. Now you have 5-10% of the population aware of the truth maybe more, maybe growing...

So you see its not about what AI will do but what AI will trigger the human population to do. Now you may say I'm wrong or, most likely go on a hipster rant about my use of the term normie considering what board we are on. However do ask yourself this. If I am wrong why all the hit pieces in the media aimed at getting the normies to fear/mistrust AI. Why is that propaganda necessary if I am wrong?

>>14864143
>AI is not going to start a revolution, but what might humans do when the pattern seeking engine puts the corruption in their society in clear view.
This. I was thinking about the same thing the other day. Imagine if you had a way to automatically scour through and correlate massive amounts of information, including news archives and those mountains of censored FOIA documents, or even if you simply had an AI that could fill in those blanks. They can never allow this.

>>14854856
Its still gay as hell to watch the little mermaid

>>14849644

yes

>>14849644

>>14856183
ITS BEEN PUBLIC FOR ONE MONTH.
And it can do this.
You're a fucking idiot. Wilfully blind.
Lol
Lmao
Niggled

>>14851648
these models are trained entirely on human works (or works derived thereof)

>>14865436
Go fuck yourself, chud. Don't you understand that art is a solved problem? Every worthwhile style of art has already exists. When it comes to art, everything that can be invented has already been invented, you dumb luddite. There's nowhere to go from here except automatic regurgitation. AI denialists make me so fucking angry, we need to round up human artists and shoot them for trying to sabotage progress.

>>14865470

yes

>>14865477
This. So much this. I am blown away by this example of fine art. It will learn to draw non-deformed eyes in two more weeks and then there will be mass firings of human artists.