🧵 Untitled Thread
Anonymous at Wed, 21 Feb 2024 19:23:09 UTC No. 975031
>there are currently attempts happening at making LLMs control Blender since direct mesh generation always turns out messy as fuck
This shit is fucking wild
Anonymous at Wed, 21 Feb 2024 20:22:06 UTC No. 975035
>>975031
Lmao aibros are such hacks
Anonymous at Wed, 21 Feb 2024 20:50:23 UTC No. 975036
I mean technically you can do everything in blender via python api and technically a LLM can write code.
But it's bullshit.
I've tried that, not in blender but asked it to design me a mechanical 3D printed part in OpenSCAD.
spoiler: I used GPT4 and it failed
Anonymous at Wed, 21 Feb 2024 21:06:25 UTC No. 975038
I don't understand why meshes and topology are apparently SUCH a huge issue for AI.
Especially if you look at low poly base meshes there's not really a ton going on. There are only so many ways to make a efficient solution for the shape at hand and everything is so fucking repetitive why can't they train AI on that shit?
Doing the right topology for a model is not "difficult" it is just fucking TEDIOUS and REPETITIOUS. To me that sounds like the picture perfect application for AI.
Anonymous at Wed, 21 Feb 2024 23:12:21 UTC No. 975047
How worried are people at Blender like Ton right now?
How worried is Side and big Auto?
Anonymous at Wed, 21 Feb 2024 23:46:43 UTC No. 975048
>>975047
Ton gave an interview in the recent Blender Today and he's not worried at all and actually did the sassy boomer thing. Basically claiming that currently available LLMs are "stealing" from artists and that Blender is working together with other open sauce communities to do their own thing with "ethical" AI.
Basically his endgame is to integrate AI in such a way that you can type in a prompt like "brickwall texture with ivy and grime" and it will generate it. Stuff like that. AI should aid the artist.
The stealing part was kinda embarrassing but otherwise he has the right idea.
Anonymous at Wed, 21 Feb 2024 23:48:10 UTC No. 975049
>>975048
To clarify: Mainly stuff like generating maps and AI trained on "allowed", freely available for training data.
Anonymous at Wed, 21 Feb 2024 23:57:45 UTC No. 975051
>>975038
Way to think about is that you have to understand what the actual shape is and what's important to humans about it before you can simplify it.
The stipulations you need to make about what shape to emit and what shape to keep is a phenomena that require you to already have solved a seemingly
much harder problem of providing human like spatial reasoning. A lowpoly mesh is one that not only captures the approximate volume of a thing it also
places emphasis on how the outline of the model represent something in our visual field that is analogous to how we experience the psychological construct of what's depicted.
An AI don't have the same evolved layered systems that make up our psychology and dictates what's important to pay attention to from a human vantage point.
It has it's own internal representation of constructs within a visual field that emerges independent of how we view the world.
Thing to remember is how we're not programming these things, we're growing/training them. We can't easily put a target to a distant capability like clean mesh re-topo without first having strong spatial reasoning emerge.
The good news is that with Sora such strong capabilities seemingly have emerged now which means clean AI driven topology is on the horizon.
The bad news is how close that makes us to be entirely obsolete for everything.
Anonymous at Thu, 22 Feb 2024 00:10:51 UTC No. 975053
>>975048
>Basically his endgame is to integrate AI in such a way that you can type in a prompt like "brickwall texture with ivy and grime" and it will generate it. Stuff like that. AI should aid the artist.
how can that be his endgame? Sora is way way past that, and thats today.
Anonymous at Thu, 22 Feb 2024 00:13:26 UTC No. 975054
>>975053
They have neither the finances nor the man power to keep up
Anonymous at Thu, 22 Feb 2024 00:15:36 UTC No. 975055
>>975054
What do you think open AI's endgame is, then?
Anonymous at Thu, 22 Feb 2024 00:17:31 UTC No. 975056
>>975055
>open
making money
Anonymous at Thu, 22 Feb 2024 00:20:39 UTC No. 975058
>>975056
what makes you say that? No one will have a job or income to pay them and many people are saying society is going to collapse worldwide because of their products
Anonymous at Thu, 22 Feb 2024 00:27:52 UTC No. 975060
>>975055
OpenAI's endgame is ASI, super intelligence as in something that can grapple with things we're just too mentally limited to engage with ourselves and extend us past our own capacity.
Altman himself talks about 'post scarcity society' and in his mind his efforts is seemingly all about how we'll get there.
In the talk with Bill Gates he ventures in about how 'This might be the last hard thing I ever do' and how he has issues of his own coming to terms with it.
Their officially stated current mission is creating AGI, build a 'general intelligence' that is capable of doing anything we can do.
Anonymous at Thu, 22 Feb 2024 00:30:23 UTC No. 975062
>>975060
You wanna hear it from the horses own mouth it's available here: https://www.youtube.com/watch?v=PkX
Anonymous at Thu, 22 Feb 2024 00:30:46 UTC No. 975063
>>975060
you can't force alignment
Anonymous at Thu, 22 Feb 2024 00:30:49 UTC No. 975065
>>975060
Ah, so he's a maniac and we should kill him.
Anonymous at Thu, 22 Feb 2024 00:46:02 UTC No. 975070
>>975063
and what i mean by that is this :
>"oh we're going to "steal" a bunch of shit and profit off of it and make all this money and then put everyone out of work and then go further and further and HOPE that our creation is going to still be on our side because just because guys.
Anonymous at Thu, 22 Feb 2024 01:04:21 UTC No. 975073
>>975063
>>975065
Maybe not, but the thing is the same is true for us.
Nietzsche was right and god is dead, but the thing is we're just to retarded on average to be left to our own devices so we need something to watch over us as our guardian god.
But since there is no such thing the only logical thing is to build one for ourselves.
We've got nukes up to our ears and ~half of voters believe in reptiles from inside the hollow earth etc.
At this point we might as well take our chances with an unaligned ASI because we're already living on borrowed time.
Yudkowsky might be right about alignment, but he might be wrong, I say it's worth a hail marry.
Sometimes you gotta roll the hard six.
Anonymous at Thu, 22 Feb 2024 01:08:40 UTC No. 975075
>>975060
He is fucking crazy. But its funny.
All that bullshit about ASI and shit, people uses AI to make porn and scam videos with Elon musk free bitcoin giveaways.
That's what he made, a porn Image generator and a cuckold chat bot
Anonymous at Thu, 22 Feb 2024 01:10:13 UTC No. 975076
>>975073
Some people are saying that if you truly are a tech bro and bellieve that the singularity is coming for and you will meet your new God and your energy concerns will be solved with solved nuclear fusion etc that you really are probably much closer to death than to being saved. You will be rolling, for sure, but it will just be a regular reroll
Anonymous at Thu, 22 Feb 2024 01:11:05 UTC No. 975077
>>975073
It will be fun to watch the world burn on my little farm in the middle of nowhere.
I will not have much time either but that shit would be fun
Anonymous at Thu, 22 Feb 2024 01:44:10 UTC No. 975085
>>975075
>a porn Image generator and a cuckold chat bot
Notice how those use cases is a reflection of us and not the technology itself.
What you critique there only amounts to looking down your nose on what you experience people around you are doing with it on average.
What should concern you is not what the AI reflect back to the median audience, half of which will have lower than 100 IQ and just wanna be entertained.
What should concern you is what it is able to provide those who probe at the edges of it's capabilities.
If you're still laughing at them in the face of what's going on today you're not paying attention to developments at all
or you're just too daft to understand the implications.
Anonymous at Thu, 22 Feb 2024 01:48:34 UTC No. 975086
>>975085
>you're just too daft to understand the implications.
the only implications are that my DCC subscriptions (hello, maxon, hello foundry) will have to drop in price substantially
Anonymous at Thu, 22 Feb 2024 01:54:04 UTC No. 975090
>>975051
I think image models are getting good enough and have enough topology wireframes in their training sets that they could literally just output an image with the ideal topology and then another stage could convert that to actual quads/patches
Anonymous at Thu, 22 Feb 2024 02:04:22 UTC No. 975093
>>975090
but /g/ is telling me that deepfakes are illegal
Anonymous at Thu, 22 Feb 2024 02:15:50 UTC No. 975094
like 95% of ai is porn, and the other 5% is techbros trying to use it to grift people one way or another.
honestly who cares. my own social media timelines are spammed with ai discussion and frankly it's getting tiring.
>random artist whines about ai
>techbro tries to own them epic style with a retarded slop image
>artist ratios them with an equally retarded comeback
>repeat ad nauseum
when does it end? who the fuck cares? this shit is so retarded. it can't do 2d. it can't do 3d. the videos are uncanny. literally the only thing it can do is regurgitate what it has been fed. when an AI can make something genuinely original then I'll be interested and mildly concerned. till then who gives a fuck
Anonymous at Thu, 22 Feb 2024 02:35:01 UTC No. 975098
>>975094
scroll down and actually watch the videos on the announcement page anon, you'll probably understand why everyone is so concerned. https://openai.com/sora
Everyone is concerned over what's on display because: it is time for everyone to be concerned about this. Going forward we will need you and everyone else to 'give a fuck'.
Anonymous at Thu, 22 Feb 2024 02:41:45 UTC No. 975099
>>975098
ive already seen all that shit. is ai forcibly ripping the pencil from my hand? if it disappears when someone stops paying the electric bill, then it's not a real threat.
Anonymous at Thu, 22 Feb 2024 02:42:37 UTC No. 975100
>>975098
Also understand how it's very concerning not mainly because it'll soon be able to generate flawless videos but rather because of the implications as stated here >>975091
It's about to concern everyone.
Anonymous at Thu, 22 Feb 2024 02:53:20 UTC No. 975101
>>975085
Ai is just a tool for Goons, nothing more
Anonymous at Thu, 22 Feb 2024 02:57:13 UTC No. 975103
>>975100
and What are these implications you keep crying about?
Anonymous at Thu, 22 Feb 2024 02:57:21 UTC No. 975104
>>975100
That post its completely wrong, it needs much more than that do do the task ge gives him
Anonymous at Thu, 22 Feb 2024 03:06:37 UTC No. 975105
>>975104
It's not wrong. The LLM's have multi-step reasoning that is dangerously close to be able to do real functional work and not just isolated tasks.
When you start stringing these systems together you end up with a synergy between those components that provide you with opportunity to
train for new capability.
What's happening today is we're watching subsystems being developed that underpin what needs to happen before we're able
to have autonomous machines out and about in the real world and act n accordance with all the complexities that takes place within the visual field there.
Anonymous at Thu, 22 Feb 2024 03:14:22 UTC No. 975107
>>975105
you're living in fantasy and you've seen too many movies.
Anonymous at Thu, 22 Feb 2024 03:57:52 UTC No. 975112
>>975107
>you've seen too many movies.
Hey, I wish you where right but this isn't sourced in fiction but comes from observations of confirmed current day events played out as lived reality.
You're not the only one who's in for a rude awakening tho, because of how there are so many like you that don't pay attention
or lack the background/curiosity to sit down and make sense of what you experience observing events thru your very own mk.1-eyeballs.
We're just not gonna pivot in time to reach a good outcome during the transition phase.
Just keep your eyes peeled as this progress and every time you see something suddenly advance beyond was possible the day before
ask yourself just how much people like me are being bullshit alarmists.
This is us pulling a fire-alarm because how we see smoke seeping in under the door-frame.
When you start sniffing said smoke for yourself please consider what's been relayed to you to contextualize events.
Anonymous at Thu, 22 Feb 2024 04:41:12 UTC No. 975117
>>975112
>but this isn't sourced in fiction but comes from observations of confirmed current day events played out as lived reality.
absolute cringe
touch grass, please
Anonymous at Thu, 22 Feb 2024 04:43:02 UTC No. 975118
>>975117
>touch grass, please
No can do, winter conditions.
Anonymous at Thu, 22 Feb 2024 07:13:49 UTC No. 975127
>>975058
>people are saying society is going to collapse worldwide because of their products
Retarded take.
They make pictures and a chatbot. Start talking about the collapse of society when we have androids walking around taking construction jobs from illegals. Personally, I'd prefer the robot.
Anonymous at Thu, 22 Feb 2024 10:30:29 UTC No. 975136
>>975112
That's so far away from reality, I understand those concepts, I just think that we are not as close as you think.
Anonymous at Thu, 22 Feb 2024 10:43:13 UTC No. 975138
>>975105
>LLM's have multi-step reasoning
They don't have any reasoning and anyone who claims otherwise is either bullshitting you (gonna keep that VC money going) or doesn't understand how LLMs work on a basic level.
Anonymous at Thu, 22 Feb 2024 11:22:03 UTC No. 975140
>>975138
Look at what happens when people throw GPT4 into a analytical loop anon.
They're very much capable of having the model current gen reflect and refine it's own output zeroing in on an answer beyond the one-fire attempt of a regular prompt.
This was the whole hoopla when people lost their shit over 'GPT agents' few months back.
People who claim LLM's can't reason probably have a very skewed or romantic view of how we ourselves are able to perform such tasks.
Nothing that happens in our own brain is magic, the digital neural tissue we've engineered altough different and simpler in architecture
is very much capable to operate in ways similar to how we work our way thru a problem.
Autonomously chain-prompted the right way they can find valid solutions to problems by challenging it's own ideas about the problem and refining it's framing til a satisfactory answer has been discovered. If that doesn't pass under your definition of 'reasoning' your definition of reasoning is incredibly narrow.
Atm you really have to be a GPT power user yourself or pay attention to the field to know about these things but once GPT5 releases in a few months
and it's both overall smarter and also have these 'GPT agents' like functionalities built in as default you'll hear everyone talk about this.
Anonymous at Thu, 22 Feb 2024 11:31:04 UTC No. 975141
>>975060
>Altman himself talks about 'post scarcity society' and in his mind his efforts is seemingly all about how we'll get there.
Keep in mind that he's selling a thing. Elon Musk does the same thing by talking about how he wants to go to Mars. There is some truth to it when these tech visionaries talk about how they're so committed to their ultimate goal, but the way they talk about it with absolute unwavering certainty is all for show, to make the investors and personality cultists happy.
Anonymous at Thu, 22 Feb 2024 12:08:54 UTC No. 975146
>>975141
Musk is a very clever man in some aspects but if his general intelligence was modeled as the surface of a sphere I imagine it would look a lot like pea with a few tall spikes sticking out of it. He's a incorrigible grifter with the psychology of a man-child prone to tantrums, desperate for attention and affection.
Details of his personal life paints the picture of this increasingly unhinged substance abuser who make rash 'spur of the moment' decisions left and right.
There's nothing in Altman's demeanor or personality I've seen so far that suggest to me the man is compromised in any way that warrants a comparison to Elon.
He comes across as extraordinary level-headed and genuine, he doesn't mince his words and tend to answer the spirit of the question people ask him
even when it's something adversarial/sensitive or a bit stupidly phrased.
He has too much power for me to ever trust him but he's of a favorable type of psychology that you want in those people who end up holding such heights of power.
If a Elon type individual was in control of OpenAI I would carry a PEZ dispenser with cyanide on my person at all times.
Anonymous at Thu, 22 Feb 2024 12:19:40 UTC No. 975149
>>975138
If you actually understood how LLMs work, midwit, you would know that nobody, not even the people working on them, even know how they think. That goes for all generative models. They are quite literally coming up with emergent capabilities before the researchers' very eyes.
Anonymous at Thu, 22 Feb 2024 12:29:19 UTC No. 975150
>>975149
I watched the computerphile videos on the subject where they explain pretty well that while it may look like it's reasoning it's really not.
An I yes trust a compsci professor on youtube more than a """industry""" CEO.
Anonymous at Thu, 22 Feb 2024 12:47:01 UTC No. 975151
>>975150
>compsci professor
Oh, but that's just the thing. Those "experts" were all wrong about their AI predictions as well. The truth is, there are missing pieces that nobody fully understands here. Reliable reasoning and spatial reason could be a single digit number of algorithmic breakthroughs or compute power magnitude increases away.
Anonymous at Thu, 22 Feb 2024 12:49:27 UTC No. 975152
>>975150
Explain how >>975143 isn't reasoning then. Is there a text somewhere on the internet about how to navigate a structure I just made up the AI could copy/paste from?
If this isn't reasoning what word would you use to describe what the AI is doing there?
Anonymous at Thu, 22 Feb 2024 13:34:48 UTC No. 975153
>>975054
Blender foundation, a million dollar project can’t hire more people to build stuff. Sounds like corporate culture 101 since we know the developer team refuses to fix certain issues.
Anonymous at Thu, 22 Feb 2024 13:53:30 UTC No. 975154
>>975149
>They are quite literally coming up with emergent capabilities before the researchers' very eyes.
you sound like you are one of those people who believe that everything is going to be "solved" for you and you are just going to vegetate and feed on a little tube while being immersed in a pod, like in the Matrix. How the hell can you _actually_ believe this shit? You're delusional. It's all fakery.
Anonymous at Thu, 22 Feb 2024 15:17:52 UTC No. 975158
>>975154
That's not what he said at all, 'matrix pod' is all you. Capabilities emerging in these networks is an objectively
correct description for how training and scaling these things works. Even ANN literature from decades ago will tell you as much if you where familiar with the field.
As for the spirit of the question of 'how can anyone think that we'll be reduced to have nothing to do.' It arises from how these things will scale.
You raise and educate a person to be Albert Einstein you have exactly one Albert Einstein.
You teach an ANN to be an Albert Einstein you instantaneously have this vast army of Albert Einstein's that can work on various problems and you can just keep
adding new ones at the same rate you can manufacture additional hardware to run more instances.
Now if you can build a robotic body suitable for working in the real-world cheaper than you can raise a human, putting it thru education, pay it's salary and all the insurances etc
how can anyone compete? They can't. These machine bodies will eventually cost way less resources and energy than say building a new car to ferry someones inefficient meat-ass around.
When you can have a replica of the best mind we currently can build remote operate that body anywhere any time and have it work 24h hours around the clock without a salary or complaints for long hours etc, that is the process by which we all end up with nothing commercially viable to do.
You follow developments you'd know how much of this technology already exist as convincing prototypes and what they're already able to do this day.
Couple this with the break-neck speed that new developments currently happen on the 'let's build a mind' front that is how you get a lot of people
taking this 'silly sci-fi sounding scenario' to outsiders as very real and very imminent.
Anonymous at Thu, 22 Feb 2024 15:32:32 UTC No. 975161
>>975158
>Now if you can build a robotic body suitable for working in the real-world cheaper than you can raise a human, putting it thru education, pay it's salary and all the insurances etc
how can anyone compete? They can't. These machine bodies will eventually cost way less resources and energy than say building a new car to ferry someones inefficient meat-ass around.
>machine bodies will eventually...
you sound completely insane. This is Y2K all over again if you are old enough to remember
Anonymous at Thu, 22 Feb 2024 15:37:35 UTC No. 975162
>>975152
I ask 10 times to change a word and never did it, so not reasoning at all, a human could understand specifically what i was asking but not the Ai.
I am not that impressed
Anonymous at Thu, 22 Feb 2024 22:14:27 UTC No. 975202
>>975161
This is totally different than Y2K because there are billions of dollars going toward ensuring that it will happen, and every step along the way can help finance or even directly contribute technologically to the endpoint.
The incredible power of a technologically advanced civilization pushing toward something tends to make it more likely to occur
Anonymous at Thu, 22 Feb 2024 22:21:20 UTC No. 975205
>>975202
nothing is going to happen. Your ai hopes are opposed to capitalism.
Anonymous at Fri, 23 Feb 2024 00:27:40 UTC No. 975220
>>975202
>>975205
> because there are billions of dollars going toward ensuring that it will happen
Not just billions, OpenAI is currently in the process of raising 7 trillion dollar for an 'AI chip manufacturing project'.
This is another aspect why some are starting to freak out they're soft-launching AGI as we speak.
Breakthroughs have happened in their labs that warrant they raise a budget that rivals like the annual budget of entire nations on something.
$7000.000.000.000 isn't something you pitch unless you have a fantastic pitch.
Anonymous at Fri, 23 Feb 2024 00:35:01 UTC No. 975223
>>975220
>Breakthroughs have happened in their labs
>they are stealing left and right off of everything they can download without giving any credit
again, open ai's goals are diametrically opposed to capitalism and won't work in a million years
Anonymous at Fri, 23 Feb 2024 02:20:56 UTC No. 975237
>>975223
>diametrically opposed to capitalism and won't work in a million years
In the longterm? Yes. In the short term? No. Do you know how much capitalism is known for planning ahead past the nearest quarter?
"Line goes brrrrp!" til it stops and then everybody is like "oh damn... we cant grow forever lolwut???"
By then it's already too late and we're in ~space-communism or we're all dead.
Capitalism isn't the end-state of civilization anon, it is failing before us as we speak.
This is why anyone with plus-sized thinker has been talking about 'late state capitalism' for at least the last decade or so.
Anonymous at Fri, 23 Feb 2024 03:01:51 UTC No. 975240
>>975237
>we're in ~space-communism or we're all dead.
We're already in space. I'm middle aged now and my life was a lot better 20 years ago. My life is over, anon
Anonymous at Fri, 23 Feb 2024 03:28:02 UTC No. 975244
>>975240
>I'm middle aged now and my life was a lot better 20 years ago.
I'm same as you anon. I'm too middle age and my life was a lot better 20 years ago. Thing is civilization is like that and it goes thru it's boom and bust cycles.
Never asked you grand parents what it was like growing up around WW2? They had some different ideas about what constitutes a shitty day.
We got to have our youth in what currently appears as this absolute golden age of human civilization.
If everything is taken away from us tomorrow - who the fuck cares, we've already won.
So bucko up samurai, it's time to eat salty popcorn and watch how the cookie crumbles.
Anonymous at Fri, 23 Feb 2024 04:08:15 UTC No. 975248
>>975244
>Never asked you grand parents what it was like growing up around WW2?
>We got to have our youth in what currently appears as this absolute golden age of human civilization.
my grandparents all died when i was young. I grew up poor and am still poor. Now I am also old
Anonymous at Fri, 23 Feb 2024 04:22:46 UTC No. 975249
>>975220
This seems to be related to the leaks about the 'Q-star' algorithm last year and other companies are also vectoring in on the same thing.
Former GitHub CEO and partners just dropped a modest 100 million dollar on a start-up that claims to have engineered the same sort of algorithm.
They state they're currently witnessing what they call 'active reasoning' emerge within their LLM's.
Seems this triggered a 'race to the bottom' in which OpenAI look to capitalize on their lead by securing -ALL FUNDING- to ensure
they're the ones who bring the new age about.
So basically get on amazon and buy a 4-point harness for that meme race-seat gaming chair of yours, because things are about to come at us real fast.
Anonymous at Fri, 23 Feb 2024 10:42:01 UTC No. 975273
>>975237
>Capitalism isn't the end-state of civilization anon
Yes it is, communism never worked, if EEUU goes to that route, it will become a bigger Venezuela, that's why they train AI with only left leaning ideas, because if they didn't, Ai would not lean left, we know that, they shut down tons of ai because of wrong thinking. Don't be retarded, Ai is not free thinking, Ai is trained to be left leaning. Just wait until a free Ai is created. That would be a fun day
Anonymous at Fri, 23 Feb 2024 11:01:59 UTC No. 975278
>>975273
It's the current end-state but history doesn't end with us anon, well hopefully anyways, with nukes in the equation you never know.
Do you seriously think if we're around in year 3000 we're still gonna be governed by populist elected government that serve to
facilitate the interface between worker citizens and private capital who's lobbying for less regulations?
That we'll have a speculative stock market and a boom-bust economy?
It'd be as retarded as a Viking inventing a flux capacitor for his longboat and travel to our current day
only to find we still go to the Jarl and request trial by combat whenever we're in disagreement with our peers.
Anonymous at Fri, 23 Feb 2024 13:34:50 UTC No. 975281
>>975278
>Do you seriously think if we're around in year 3000
Get real anon, you are appealing to one thousand years from now. Get real
Anonymous at Fri, 23 Feb 2024 16:17:04 UTC No. 975291
>>975060
>>975073
It would be better to genetically engineering ourselves into being better than to try and create some utopia with a super advanced AI.
The problem the tech bros have is they are hyper fixated on this utopia they aren't putting up proper guardrails on this technology. You want to create an AI god to solve all of our problems but this is more likely to backfire.
Anonymous at Fri, 23 Feb 2024 16:27:20 UTC No. 975292
>>975146
Altman is worse than Elon. Elon is a greedy grifter, Altman is a utopian thinking he's doing humanity a favor. He pursues his goal with religious like zeal. A utopian is far more dangerous than a grifter, because in the mind of the utopian a lot of terrible things can be justified in the name of their goal. Most of the team is like this, this is why they didn't get permission to use copyrighted images. They admitted that their goal is so important that they would rather ask for forgiveness than permission.
Anonymous at Fri, 23 Feb 2024 16:32:31 UTC No. 975293
>>975281
Well If we start by having you admit we likely won't live in the same society in 1000years you admit how it's not the end state as you initially claimed, then we can obv start walking it back from there to pin down what our actual disagreement is.
1000 years is just a figure I throw out there to point out how unreasonable claims capitalism can't be superseded by something else ultimately are.
I myself ofc expect it to start happen a lot sooner than that.
Anonymous at Fri, 23 Feb 2024 18:01:33 UTC No. 975302
>>975278
No, I never said anything about government ,future is without government, no regulations of any kind, just a free market society.
> It'd be as retarded as a Viking inventing a flux capacitor for his longboat and travel to our current day
only to find we still go to the Jarl and request trial by combat whenever we're in disagreement with our peers
Yeah that would be stupid, but we still do it when it comes to countries no? What about Russia and Ukraine, I don't have a fucking utopic machine that solves government wars? If you gave one please send it to Putin or Zelenskyy so no more Slavic people get killed.
Utopias don't exist, grow up, that's childish. A lot of people got killed because of those retarded ideas.
Anonymous at Fri, 23 Feb 2024 19:10:47 UTC No. 975314
>>975302
>future is without government, no regulations of any kind, just a free market society.
Without government there is no society and no freedom anon. The idea of agreeing on what the rules are, that there is at least theoretically 'equality before the law'
and having an institution with the capability of enforcing those laws is the foundation capitalism and the free market rests upon.
Remove government and remove all regulations you'd have a perpetual race-to-the-bottom with 'might is right' as the only instrument to dictate outcomes on the playing board.
Corporations unchecked would behave like the robber barons of the dark ages after the fall of the Roman Empire where you venture out of your heavily armed fortress to rob
at swordpoint whatever shit you can lay your hands on before returning back to base to divide the spoils amongst your vassals.
I'm giggling like a little girl here how you propose this cookie-cutter super obv 'darkfuture cyberpunkesque' type scenario as your proposition
and still manage to call something 'childish' at the end there of your insane post after flying off in a tangent about Ukraine.
Anonymous at Fri, 23 Feb 2024 20:03:45 UTC No. 975319
>>975314
>Without government there is no society and no freedom anon.
I can do exactly the same institutions based on a private sector without any type of government existence. Private security for example is exactly based on free market principles, you don't pay shitty security, and shitty private security goes out of business, i can't say the same about police.
Corporations are people,and they are tied to the same thing that makes this society possible, free market.
A corporation is not tied to government force, neither they work like governments. They only serve to give goods cheap to make money. If you think that McDonald's will become a sort of paramilitary organization you are out of your mind .
I mean you are giggling like a little girl but you can't see how stupid it to believe that the government regulating your way of living is freedom.
Also the Ukraine thing was not a tangent, you bring it up
I quote
>Only to find we still go to the Jarl and request trial by combat whenever we're in disagreement with our peers.
I just answered in a bigger scale. Maybe you don't need to fight because your neighbours dog shit your backyard, but you definitely have to if you government has a conflict with another.
Anonymous at Fri, 23 Feb 2024 20:27:29 UTC No. 975321
>>975319
>Corporations are people
How come you're still typing and not long dead in ~Jonestown? You drank _all_ the Kool-aid anon.
Anonymous at Fri, 23 Feb 2024 21:41:05 UTC No. 975333
>>975321
I drank the kool aid or maybe you need the government because they give you your unemployment check
Anonymous at Sat, 16 Mar 2024 15:05:46 UTC No. 977863
>>975048
>My prompts!
AHAHAHAHA I can not wait for AI regulation to hit prompters hard.
>t like "brickwall texture with ivy and grime"
You can do that into google and get a pre made asset or a generator of that thing.
What you lazy fucks do not understand is that AI is theft and will only be able to """""generate""""" the same asset it did steal.
>b-b-b- what if I steal millions of 3D models
1) There are not that many 3D models
2) It is literally more retarded then using a generator, whatever happened to the art of 3D model generators?
3) WHY? At this point it is better to sell a asset pack of all these 3D models not fucken prompt around to show the same stolen 3D models.
And 3D will be a bitch for AI. Unlike with 2D fucking up the geometry will fuck up your models.
Anonymous at Sat, 16 Mar 2024 15:08:03 UTC No. 977864
>>975060
>OpenAI's endgame is ASI, super intelligence as in something that can grapple with things we're just too mentally limited to engage with ourselves and extend us past our own capacity.
AHAHAHAHA imagine believing this bullshit after you have only written a stolen picture database with extra steps.
There is no such thing as super intelligence AI. You talking to chat GPT to only realize it is literally nonsense lies should make you realize this.
Anonymous at Sat, 16 Mar 2024 16:18:59 UTC No. 977876
>>975152
>Explain how >>975143 isn't reasoning then. Is there a text somewhere on the internet about how to navigate a structure I just made up the AI could copy/paste from?
You answered your own question.
>Is there a text somewhere on the internet
This.
> isn't reasoning then.
Because it is not. It literally can not think. It is like using a stupid search engine from the 90s. IT is not thinking.
Anonymous at Sat, 16 Mar 2024 19:20:54 UTC No. 977899
>>977876
>This.
No such text exist, how could it? It understood a situation I made up on the fly.
>Because it is not. It literally can not think.
Yeah it can't think... it just looks like it thinks...
Much like it literally can't take your job, it'll just looks like it's doing your job.
and how it wont literally rule society, it'll just looks like it's ruling society.
>It is like using a stupid search engine from the 90s. IT is not thinking.
Oh yeah? When's the last time you saw altavista do this anon:
>>977569
https://www.youtube.com/watch?v=Sq1
Wake the fuck up samuraianon. You know that Rage Against the Machine song where they screamed for someone to wake up? That's you they're singing about. Waky paky.
There's this machine to actually rage against, unplug that head from the sand pretty plox.
Anonymous at Sat, 16 Mar 2024 21:51:51 UTC No. 977912
>>977899
>understood a situation
I literally does not.
You simply assumed like some brain dead child that it dose. Here is you problem.
>Yeah it can think
Evidence?
>>977899
>Much like it literally can't take your job, it'll just looks like it's doing your job.
>and how it wont literally rule society, it'll just looks like it's ruling society.
What is this schizophrenia?
Look kid AI art is literally a database with pictures the AI uses to bring out the same pictures that the mega corporation making it stole from all over the internet.
AI is terrifying since you literally do not need more then that repetitive behavior to make a self driving car, once you feed it enough data how a road works and looks it can navigate it.
Same for every other AI task.
AI however is dumb as a rock, it has the mental capacity of a plant.
>Oh yeah? When's the last time you saw altavista do this anon:
Are you fucken retarded? You did feed it txt and it have you txt. Now you magically switch to robot grabbing objects?
>You can talk to it!
And? It is unbelievable how stupid you AI fagots are.
There is nothing magical about the robot knowing where the dishes are.
Wow feed it enough pictures of plates and it can recognize plates. No idea why voice interface is even interesting since speech recognition was something since the 90s and before. You can query the robot OS directly about the data where the plates are.
>You know that Rage Against the Machine
Now you are ranting about a song you schizo?
>Yeah it can think
OK ask it to finish computer code. And 9/10 it will fail hard.
Because it can not think. It does not think. You are like a retard who thinks a speaking toy is alive and thinking because it is repeating pre recorded voice lines given to it. That think can not think.
1/2
Anonymous at Sat, 16 Mar 2024 21:52:54 UTC No. 977913
>>977899
>>977912
2/2
> It understood a situation I made up on the fly.
You debunked yourself.
#1 Prove that no one in the history of the internet ever wrote what you have written.
I wait.
>I le typed it so no one else ever did type it
AHAHAHAHAHA.
#2 Prove that on one in the history of the internet ever did write anything about. Something as generic as a floor numbered from 1 to 9
Jesus fuck you are stupid.
If you know what to ask a AI you can see it breaking down and spit garbage.
Anonymous at Sun, 17 Mar 2024 04:46:01 UTC No. 977937
>>977912
>. Now you magically switch to robot grabbing objects?
Are you that dense? That there is the yet to be released version of GPT combined with robotics from Figure and the same tech behind Sora for interpretation of the visual field to serve as it's eyes. It's the very same type of mind running in that thing that runs behind chatGPT.
These things understand context very well.
>OK ask it to finish computer code. And 9/10 it will fail hard.
It's getting scary good at that too, people have been putting out loads of demos of what they're able to do with autonomous AI agent 'Devin' just this week it's capable of autonomously completing sophisticated tasks. A 100 million jobs in programming is immediately threatened as of ~middle of last week.
It's already 'gonna change society forever' level of serious and there's no sign of slowing down, developments are still accelerating.
You probably don't follow this field so you have no idea of far along we are and what already exit today that you'll hear about in mainstream news soon.
Everyone should know what's going on because the world is changing by the day.
This post is intentionally alarmist because the situation is now such that it is legit alarming.
Anonymous at Sun, 17 Mar 2024 06:05:19 UTC No. 977948
>>977937
>These things understand context very well.
You have schizophrenia.
> That there is the yet to be released version of GPT combined with robotics from Figure and the same tech [SCHIZO RAMBLINGS]
>It's getting scary good at that too
It actually does not. And this is the point! Robots can do repetitive tasks like garbing objects or detecting faces since the 1000s of stolen pictures and videos of human faces that the mega corporation stuffed into the AI are basically the same for all humans at this point or all plates.
Yes these are stolen videos I do not give a shit what the I consent form was on youtube and the TOS so google can now use your voice and video you uploaded to youtube.com to train its AI.
>jobs in programming
Look schizo the subject was if the AI can think not if it can take jobs. You can not understand this subject.
>jobs in programming
Fun fact most of it is stolen code.
>AI write a hello world program
>>Sure thing let me copy past this stolen code that is in my database.
>AI write me a connection to unity game engine API
>>Sure thing let me copy past this stolen code that is in my database.
>AI write this driver for this hardware that no one has ever written
>>[literally nonsense garbage]
>It's already 'gonna change society forever' level of serious
I know and human society is over since ROBOTS and AUTOMATION have arived to this point. You can look up old predictions from the 1950 and 1960s how AUTOMATION and ROBOTS will take jobs and are a threat to society.
WE did know this for a long time.
The only problem here is that corporation will replace humans with chat GPT powered talking people over the phone.
Robots will take 99% of all jobs. This is the dangerous part.
However this is a completely different subject from what you are rambling all about
>My talk talk toy is alive and super intelligent because it answered me!!!!!!!!!!!!!!
This is the problem here schizo. You keep changing subjects.
1/2
Anonymous at Sun, 17 Mar 2024 06:06:23 UTC No. 977949
>>977948
>>977937
2/2
AI can not think it is dumb as a rock, basically plant intelligence. HOWEVER you do not need more then plant intelligence to replace all human phone operators. The robot you demonstrated is can not think it is basically a
>Press 1 if you want to talk about subscribing to our service
>Press 2 to register a complaint
>Press 3 to talk to legal.
Nothing more
IT
CAN
NOT
THINK
You schizo fuck!
>This post is intentionally alarmist because the situation is now such that it is legit alarming.
Yep. Only the AI is not some ScienceFictin faggotry super intelligence or some shit it is a non thinking dumb as a rock neural network. In reality only the super rich will benefit from this by replacing human workers and making more money.
And the subject is blender and 3D not talking on the phone chat bots. And AIs can not into 3D because 3D is demanding.
Before AIs threaten 3D they will be regulated as fuck since all the other professions will complain about them to the government.
Anonymous at Sun, 17 Mar 2024 06:29:54 UTC No. 977953
>>977949
>IT
>CAN
>NOT
>THINK
Tell me how it is you fell so ALLCAPS confident in this anon?
You've seen it carry out long and complicated task successfully. How can you explain the model succeeding at this if it isn't reasoning?
You know it's hardware is modeled by studying the mechanics of biological structures we know first hand is able to understand things and reason about things.
When you see the machine brain we built succeed in tasks we train it for by processes that are analogous to those we've observed to take place in biology.
How can you be so 100% sure it isn't reasoning?
We designed ANN's for a long tome with the clear intent to understand and reason. Now given recent advancements in the neural architecture that underpin these
modern variations it succeed at clears tests at ever more complex reasoning we throw at it demonstrating it works.
Witnessing this, knowing the origin and intent of the technology how can you be so absolutely cocksure it 'can not think' ?
How do you envision it is possible for you and I to able to sit here thinking if not thru all those stored weights in our neural hardware?
Anonymous at Sun, 17 Mar 2024 08:44:54 UTC No. 977965
>>977953
>Tell me how it is you fell so ALLCAPS confident in this anon?
Because know the technology and I know how to make it break.
>You've seen it carry out long and complicated task successfully
WTF are you talking about? Define "complicated tasks"
>[trans humanist fart sniffing]
Because I unlike you understand how the technology works.
You are some retard who obsesses over transhumanist garbage. You jump from
>Wow it can answer and talk about stairs!
I challenged you you retarded schizo to prove that literally no one wrote on the internet about something as trivial as stairs and walking them up to rooms. You failed this.
Want to see a AI break?
Talk to it about a game you make up on the spot and then play this game with the AI it will be unable to do it.
Because it does not understand the words it can not think it is reacting to what you type and it is reacting based on normal simplistic human responses.
Small talk is trivial and everyone can simply write down a long chain of responses to common phrases
>Hello how are you
>>>Good
>You?
>confident in this anon?
OK let me explain this in all caps to you
BECAUSE IT CAN NOT THINK!
Do you even understand what thinking is?
It is a non thinking prompt who reacts to your TXT input. It has no goals.
The moment when it starts thinking is the moment it starts on its own spiting out schematics for technology it came up with.
Oh and without the AI corporation stuffing billions of stolen human work into it.
It can not think at best it is a search engine and at worse it is photo bashing things in it into something.
>confident in this anon?
Because of the evidence?
AI repeatedly will make up things and tell """""lies""""".
However these are not lies, it can not understated truth VS lie it is only photo bashing what was fed into it.
https://www.youtube.com/watch?v=Gmp
https://www.youtube.com/watch?v=oqS
PS:
>>>I am a big bing!
Anonymous at Sun, 17 Mar 2024 09:42:41 UTC No. 977970
>he thinks the LLMs they put out to the public aren't severely dumbed down
lol, lmao even
Anonymous at Sun, 17 Mar 2024 09:44:21 UTC No. 977971
>>977970
bbbro trust me the TECH is there its jjjust in jeffs bezos' closet
Anonymous at Sun, 17 Mar 2024 10:33:48 UTC No. 977976
>>977971
an unrestricted research version of gpt4 dubbed Eureka was already used to engineer better, more novel ML rewards for robots than humans engineers could conceive, that was six months ago
only midwits like yourself don't understand the implications of this
Anonymous at Sun, 17 Mar 2024 11:58:15 UTC No. 977979
>>977965
All I see is you making appeals to how it's not already a functional human level AGI that can do anything you can do.
You hyper focus on all the ways you can have it fail if you attempt to trip it up, ignoring what it already enable people who want it to succeed
have figured out how to have it solve.
Is that really your cope anon? It came from nowhere last year and the research variants are already doing tremendous things.
Your cope is that the public facing chatGPT variant isn't already infallible and therefore strongly declare it's unable to reason/think/reflect etc.
You don't find it at all telling how a crude early form of true AI like chatGPT 3.5 is already able to conduct these simple reasoning tasks running on a
neural net that have neural pathways of roughly ratbrain size?
If a machine ratbrain is spitting out the sort of answers we get from GPT 3.5 what do you suppose is gonna happen when they scale this and train something with the
100 trillion connections of human complexity? You're not at all worried how we're seeing sparks of humanlike intelligence even in something this simplistic?
To me you're like someone screaming at a rat for not solving calculus and playing complex games with you out of the box
instead of marvel how it can talk to you coherently about difficult subjects and help you draw conclusions you otherwise wouldn't reach.
Saying that it can think/understand and reason doesn't stipulate we think it has human equivalent cognitive abilities sentience/qualia etc.
It's a type of intelligence completely alien to our experience but it's still intelligence.
Anonymous at Sun, 17 Mar 2024 12:13:18 UTC No. 977980
>>977965
A jumping spider won't play tic-tac-toe with you either but if you observe it you will notice how it processes it's environment and execute coherent actions.
Is there a qualia associated with being a jumping spider? Who's to say, nobody know how complex animals need to be for things like that to emerge. What's evident is that it's clearly intelligent and capable of certain types of reasoning.
Perhaps your understanding of words like 'intelligence' 'reasoning' 'sentience' etc is too naive/narrow to grasp what people are getting at in saying GPT understands things.
Even a bee or and ant understands things and few think they are more than bio-robots following scriptlike behaviors.
chatGPT is not insect levels of complex tho, it's more analogous to a small mammal in it's architectural complexity.
Is there qualia associated with being a cat or dog? I'd say most def. Rat? Watching my friends pet rats I have a hard time convincing myself no-ones there, they seem very sentient.
Doesn't mean chatGPT has any, it didn't evolve in a world where it needed any agency to survive as everything it needed to come about was handed to it.
But it's not ruled out even sentience could emerge on these levels we're currently probing provided what we witness in biology.
But do understand that even a 'Philosophical Zombie' type entity unawake and unaware can be 'intelligent'.
Anonymous at Tue, 19 Mar 2024 10:25:39 UTC No. 978155
>>977970
I see cope and seethe over me disproving you.
AI is dumb as a plant it is reactive it reacts to your actions based on the most likely things other HUMANS have written in all the data that was STOLEN from the internet.
These AIs are dumb as a rock, their photo bashing way of thinking is interesting however not real intelligence. They do not understand the concept of a lie or even truth. They will make up things and it is not like the early AI worshipers liked to cope
>LE it is joking
No it is not the AI did go crazy and is making up words. You are not taking to something that can even think. AIs are literally closer tho wind up toys who play prerecorded messages then anything alive.
Even the word AI is cringe as fuck it is artificial however it is not intelligence. It is only a neural network.
This doese not mean AIs/NNs are not impresive nad dangerous they are only not in the
>OH MY GOD IT GET SUPER INTELLIGENCE AND IMPROVE ITSELF AND CREATE THE SINGULARITY
More in the
>Oh shit the corporation can replace me with a robot who uses a AI/NN who was fed billions of data to do my repetitive job.
>Muh LLMs
are you fucken retarded?
How are large language model even a thing? You keep spamming this garbage however over 80% of all robots do not need to understand langege you dumb fuck!
Why the fuck does a automated AI turret gun need to even understand any language?
I literally does
1) Detect target
2) Point at target
3) Shoot.
It literally does not need more, same is true for about 100% of all factory robots.
THEY
DO
NOT
NEED
A
LLM!
You biggest retard on the planet!!!!!!!!!
A factory robot does not need to understand your words! Or even fucken speak! In reality NNs are used inside a traditional hard logic OS. The NN is used for tasks like detecting a bad apple(fruit) to be rejected.
1/4
Anonymous at Tue, 19 Mar 2024 10:26:41 UTC No. 978156
>>977979
>>977980
>>978155
You will use 2 NNs together one to do the rejecting and another to detect if the robot failed to throw away the apple. The detector can then literally flip a electric switch for a light and buzzer to sound so the human operators come to see where the robot fucked up. Or since it is all software you can literally trigger anything with this from pre recorded MP3 files to speak in human language or sending SMS or e-mail to the operators that Machine #5 fucked up. There is no LLM in this you retard!
Because it is not needed.
And traditional computer GUIs are superior to you trying to verbally tell the machine
>OK robot adjust the speed to +0.45
Especially in a noisy factory environment.
Same for words like
>STOP
Better to fucken have a physical button for this or a remote signal to deactivate the robot OUTSIDE the NN!!!!!!
Another reason why all of the NN are under a master OS that is written in a traditional(HARD LOGIC) way.
If the NN makes the robot go crazy and throw stuff around the factory better to use a hard logic command do deactivate it.
And as for using verbal commands these NNs will be used for home use mostly and a novelty.
The fact that you think LLMs are interesting shows how clueless you are. LLM will NEVER lead to super intelligence or cringe shit like this. LLMs can not even lead to a NN learning to walk. And making a robot walk by itself is a far more frightening and superior use of NNs then playing
>Predict my next word
With a useless NN!
You can equally get excited over the "predict when I will take my next shit" NN or the "predict when I will orgasm NN" however even these made up examples are billions of light years above(more frightening in their use) the useless shit that LLMs are!
2/4
Anonymous at Tue, 19 Mar 2024 10:27:42 UTC No. 978157
>>977979
>>977980
>>978156
>AGI
Fun fact
AGI is literally intelligence all the things AI fagots say are intelligence is not intelligence.
>It came from nowhere last year
What did? last year s 2022. And Alexa was on the market before this.
I'm actually curious where you think useless shit like LLMs can be used? Since LLMs are nothing more then mirror for parrot 2.0 where the stupid bird thinks there is another friend on the other side of the mirror.
I see the damage that LLMs will do to society once corporations start to sell AI companions and GFs who lonely no life NEETS and other autists can use for companionship the same way you buy a mirror for your parrot to give it a friend. However that is the only use for LLMs literally!
Since the rest like stealing the work of human writers will be ended with a lawsuit. LLMs have no use outside of this.
> already able to conduct these simple reasoning tasks running on a
Name this or it did not happen.
>My magic reasoning tasks
Like? Once more you are getting excited over a toy that plays pre recorded MP3s! It can not fucken think because this is not how any of this works. I has no goals or anything. It literally responds to your prompts! It does not think because it does not think.
No prompt for it to respon = literally no action.
>neural net that have neural pathways of roughly ratbrain size?
Bzzz retard Bzzz Wrong.
The size of a LLM is fucken irrelevant the same way that the size of a whale brain is irrelevant. The NN was trained to predict what words come next. This is it. It can not escape this constraint. LLMs are useless. You have no fucken idea how this technology works.
>of humanlike intelligence even in something this simplistic?
There is literally ZERO intelligence. In like none!
3/4
Anonymous at Tue, 19 Mar 2024 10:45:11 UTC No. 978158
>>977979
>>977980
>>978157
What part do you not understand that chat GPT has no fucken goals and only reacts to prompts? Even animals have goals and will do things on their own. LLMs do not. Stop prompting it and see what will happen. There is no thinking only react to prompt program.
>To me you're like someone screaming at a rat
A rat can move and will do things on its own like digging a hole and looking for food, it has goals. a LLM literally does not and can not have. It literally was trained to predict what the next word will be. It only reacts to prompts then does nothing.
It will not turn magically into terminator. Your entire """"""knowledge""""" comes from fictional TV and singularity fags.
>talk to you coherently about difficult subjects
Once more you literally bird brain in the body of a human. You are like a bird not understanding that this is a mirror and there is no other bird next to you. Like I did say human small talk is very simplistic. And it having the right small talk points for you is literally not impressive and not intelligence.
For real machine intelligence we need to wait 3 or 5 more paradigm shifts.
1) Basic machines and 1700s clock words
2) Electric motors of the 1930s
3) Computers and programming.
4) OSs and MULTPLE programs running under them
+1) ????
+2) ?????
+3) ?????
+4) ???
+5) ?
?6?) Actually thinking machines.
NNs are a distraction and a road with no future at best.
The real paradigm shifts start once humans will scan organic brains into computers. OR even into new devices something like FPGAs only for brain cells. TO actually see how animal and humans think.
NNs are a distraction at best, does not stop it being a totally disruptive technology to society.
4/5
Anonymous at Tue, 19 Mar 2024 10:46:14 UTC No. 978160
>>978158
>>977979
>>977980
>qualia
Qualia is literally bullshit that can not exist.
Qualia is a nonsense word.
>but it's still intelligence.
It literally is not.
>but if you observe it you will notice how it processes it's environment and execute coherent actions.
And this is something a LLM will literally never do!!!!!
It literally reacts to your prompts you dumb fuck!
How stupid can you be?
>A jumping spider won't play tic-tac-toe with you eithe
YES however a spider is an animal and is non intelligent (only MOST humans are intelligent no other animal).
>sentience
Another nonsense word.
>bee or and ant
A bee or and ant are literally bio robots following scriptlike behaviors.
Are you Quolia fags denying DNA and evolution now?
> scriptlike behaviors.
However this is literally what animals are. How the fuck do you think a queen bee knows what do do once she gets pregnant? IT is programmed in her by evolution. Same for most animals.
However ants and bees have goals LLMs can not. LLMs literally react to your prompts and only to your prompts. If you do not prompt then a LMM is useless.
>qualia
>sentient
Look the retard uses nonsense words.
>My pat rat
LOL it is literally a bio robot. IT has no intelligence.
>No one is there
More nonsense words from the qualia fagot. All animals are nothing more then bio robots they have goals and act according to their evolved nature. You failing to see the limitations in their bio-programming shows how retarded you are.
>My pat rat
>qualia
>sentient
Yea and other retards got obsessed and think their literally factory made Sony toys are alive and have souls or some shit, see sony robot dog Aibo.
Humans like you are empathetic meaning you are very retarded and submissive seeing cure things in everything.
5/6
Anonymous at Tue, 19 Mar 2024 10:47:15 UTC No. 978161
>>978160
>>977979
>>977980
>it did evolve
Are you denying reality now?
Chat GPT did not evolve! IT was made artificially by humans to predict the next word. Like a fucken plant who evolved to grow into the sun.
Only chat GPT never evolved it was made by fake evolution in the computer to predict the next word. It literally can never get above this BECAUSE THIS IS HOW ARTIFICIALLY MAKING A LLM WORKS YOU TOTAL RETARD!
6/6
Anonymous at Tue, 19 Mar 2024 11:56:02 UTC No. 978173
This AI hype cycle is over. See you again in another 20 years.
Anonymous at Tue, 19 Mar 2024 12:59:35 UTC No. 978176
>>978173
Uhm.. have you seen what happened just this week and last anon?
NVIDIA's presentation, figure01, what Altman relayed in the new Lex Fridman interview?
We're already going fast and if the rumblings about what GPT5 will be able to do we're about to hit the warpdrive.
If this train has any brakes we're yet to see the signs of them.
Anonymous at Tue, 19 Mar 2024 13:35:41 UTC No. 978178
>>978176
>have you seen what happened
No. Because I don't care. Because this hype cycle for AI is over.
Anonymous at Tue, 19 Mar 2024 15:36:15 UTC No. 978184
Anonymous at Tue, 19 Mar 2024 15:49:48 UTC No. 978187
>>978178
>>978184
Delusional. You not liking AI is absolutely fine, you rejecting reality and inventing your own is just beyond ridiculous.
None of this will go away because you close your eyes and plug your ears.
>"This industrial revolution is over!" "what do you mean Franz, they're installing the steam engine and firing people as we speak?"
>"It's over! lalalala"
Like you're trying to will something out of existence by shitposting on 4chan.
Anonymous at Tue, 19 Mar 2024 16:03:53 UTC No. 978190
>>978187
Dear Sam Altman Fried, why are you shitposting about AI on /3/?
Anonymous at Tue, 19 Mar 2024 16:05:39 UTC No. 978191
>>978187
Yeah bro ai is totally not a fad that's why you niggers spend your whole day shilling it on imageboards
Anonymous at Tue, 19 Mar 2024 17:44:08 UTC No. 978198
>>975048
He isn't wrong about the stealing part
Anonymous at Tue, 19 Mar 2024 17:50:46 UTC No. 978200
>>975223
>diametrically opposed to capitalism
Are you fucking retarded? What are the layoffs and corpos cutting corners in any possible way using ai doing then?
Anonymous at Tue, 19 Mar 2024 17:53:56 UTC No. 978201
>>978191
No, we talk about it here because we artist are amongst the front-line occupation to be done away with as commercially viable as a result of these advancements.
I wouldn't call worrying about the future of all human effort or ones 'p(doom)' as 'shilling' anon.
We now live in a timeline where we're gonna live thru the forth industrial revolution and face a world completely different to the one we live in today.
It can end in either utopia or dystopia compared to today but the transition will be very painful for a lot of people including anyone who make their living doing artwork.
You don't wanna hear it because how you don't want it to happen or don't believe it's happening
but what you and others like you believe doesn't matter; because it is happening.
Anonymous at Tue, 19 Mar 2024 17:57:40 UTC No. 978202
>>978201
Nta but artists rarely made money off art, and when employed where severely exploited. I was watching few artists breakdowns of how they actually make money and its often just being an influencer on YouTube/tiktok garnering sponsorships and directing to an online shop. I think this is the last avenue where artists can make money, unless you're hired to do traditional art like painting murals or sculptures.
Anonymous at Tue, 19 Mar 2024 18:56:18 UTC No. 978203
>>978201
man thats so cute what a wall of text too bad I won't read it, stay shilling aibro I respect the grind
Anonymous at Tue, 19 Mar 2024 18:59:02 UTC No. 978204
>>978176
>We're already going fast and if the rumblings about what GPT5 will be able to do we're about to hit the warpdrive.
not him, but altman is living in a fantasy land, sheer, pure fantasy.
Anonymous at Wed, 20 Mar 2024 05:38:10 UTC No. 978236
>>978203
>wall of text
>6 lines
>attentionSpan.jpg
Kids are truly getting stupider by the hour.
>>978204
That he is, problem is how we're all about to live in that same place.
You payed any attention to last week you'd know what's already ensured to happen and that wasn't even the big reveal (GPT5/Q*).
People who keep up to date with developments sounds fucking crazy because what's around the corner is fucking crazy.
The time-lines for when to expect things have been so off and everything keep arriving so much sooner than even optimists thought it would.
Tech-space people who look at what's happening now and extrapolate understand the jig is up if this continuous even for ~2 more minutes.
We're fast approaching the threshold when the average person will have their 'oh shit' moment internalizing how they're obsolete real soon.
People like me are talking to people about this stuff who don't want to hear it precisely because they don't care/want this information.
The psychological impact this roll-out will have on society is gonna get extra ugly because of how many will be caught sleeping.
Clever people struggle to make sense of their own read on this, normal and dumb people are about to process the same information and experience a range of emotions.
If some 'schizo anon on /3/' told them how a chupacabra was tearing up their backyard at least they understand what it is they're looking at when they walk outside
and spot a chupacabra tearing up their backyard. They'll be like "oh yeah, that's just the chupacabra anon told me about".
Anonymous at Wed, 20 Mar 2024 06:22:40 UTC No. 978238
>>978236
Are you doing this for page ranking on google or something? Nobody cares. AI money has already changed hands. If you haven't abandoned ship yet, you got shafted and there's nothing you can do about it. That's all you need to know.
Anonymous at Wed, 20 Mar 2024 06:45:40 UTC No. 978239
>>978238
My value system is inherently intrinsic, money or status have never been a motivator for me in anything I've engaged with.
What's happening with this is important on a scale that should eclipse anyone's individual considerations, it's species wide.
Anonymous at Wed, 20 Mar 2024 08:35:34 UTC No. 978247
>>978239
Bwahahahaha, oh dear, a schizo
Anonymous at Wed, 20 Mar 2024 08:59:58 UTC No. 978248
>>978247
Schizophrenia is a severe mental condition anon, you're just employing it as a label to stick onto anyone you disagree with so you can make light of them.
If you reduce such a term to a label you stick onto anyone who doesn't share your general outlook on things it no longer means anything.
It's like if I start calling you a retard. You're obv not a medical retard, I'd just telegraph how think of you as someone mentally lightweight but to what end?
If I started calling you by the names I felt you deserved how would that benefit our exchange? What would anyone gain from knowing how adversarial I felt towards you?
Rise above your childish inclinations anon. We can disagree without being all retarded about it.
Anonymous at Wed, 20 Mar 2024 10:25:14 UTC No. 978252
>the schizo redditor is trying to talk to me
Eww gross
Anonymous at Wed, 20 Mar 2024 11:26:11 UTC No. 978255
>>978252
Now you attached 'redditor' as a further qualifier to lump me together with people you presumably find unworthy of your attention.
Can't you see what sort of valueless self-indulgent nonsense you're conducting here anon? You can be better than that.
Anonymous at Wed, 20 Mar 2024 11:45:17 UTC No. 978256
Btw, how is it a lot of these super hostile types always seem to know what a 'redditor' sounds like?
Are they people who got banned from that site and now perma-salty about it?
It's been numerous years now and I never understood it. Anytime I find myself on reddit it's because I goggled something where someone has typed up a relevant answer.
Seems like a pretty helpful community from the outside looking in. How do you get angry at fucking reddit unless you go actively looking for infuriating threads or something?
It doesn't make sense how these people seem to know so much about things they dislike, trannies/jannies/reddit/dialation etc.
These are power structures and activities a normal person ~never comes into contact with in their life.
Can someone give me TLDR of what archetypal 'redditor' is and why I would fit in there?
Are all these angry /pol/ sounding anons closet cases or something? Who the fuck study enough gendershit to even know half of the terms they use?
Anonymous at Wed, 20 Mar 2024 12:09:31 UTC No. 978258
>>978236
>That he is, problem is how we're all about to live in that same place.
no, everything is going to be exactly the same tomorrow as it is today. Bet on it.
Anonymous at Wed, 20 Mar 2024 13:04:08 UTC No. 978262
>>978255
>>978256
Bro got so offended he is still yapping lmao
Anonymous at Wed, 20 Mar 2024 13:18:34 UTC No. 978263
Yeah dude when QStar* comes out everyone is going to be handed a new stolen google search 2.0 and everything is going to be alright. Your wage job? You can quit it because you have STOLEN GOOGLE SEARCH 2.0. Your mortgage or apartment payment? Well you don't have a job now so you wont be paying it so your house will belong to the bank now. Your health insurance? Yeah, you don't have any now because you don't live in the EU. Food? Yeah you don't have any food now either because you quit your wage. All because you went absolutely crazy over STOLEN GOOGLE SEARCH 2.0
Anonymous at Wed, 20 Mar 2024 13:58:17 UTC No. 978267
>>978262
This is a online forum anon, if you don't wish people to post on it why do you even read posts and type responses?
Do you honestly believe I think you find my tone so hilariously offended that it has `now caused you to laugh so hard that your ass came loose?
>I have nothing to say but I will point out he can type and end this sentence in 'lmao' to let him know how entertained I am by letters I don't like reading.
Ok my dude, I'm sure everyone you debate feel like you made a strong point when you drop this novel 4D chess move on them.
Anonymous at Wed, 20 Mar 2024 14:15:20 UTC No. 978268
>>978263
This might be part of the problem why it's difficult to convey what GPT does, it requires some ingenuity on part of the prompter to have it deliver something useful. Stupid people won't understand how to use the tool because asking the right type of question is such a big part of it, so they are stuck thinking it's 'google 2.0' while those who are primed to utilize it access the deeper functionality.
Increased intelligence of the system itself might remedy it tho since a smarter GPT will be able to understand more of the context of what a stupid user is prompting for.
After all you get a lot better answers out of GPT if you articulate your question in exact and nuanced language.
It's like we're gonna see a ever larger schism form between adept language users who soak up information like sponges and those who lack mental bandwidth and/or curiosity.
Quality of ones natural language usage might be a factor that play a huge role going forward.
Anonymous at Wed, 20 Mar 2024 15:43:39 UTC No. 978269
yawnnnn
Anonymous at Wed, 20 Mar 2024 16:00:47 UTC No. 978271
>>978268
Valid point. It's pretty obv why some people ITT have such a different outlook on the value of LLM's.
Some type out long paragraphs and testify how they love it, others laughs at people for typing and dislike it. Go figure what's going on there.
The intellectual gap will widen by a lot.
Anonymous at Wed, 20 Mar 2024 16:55:34 UTC No. 978272
>>978271
Intellectuals don't spend their free time arguing on 4chan, son.
Anonymous at Wed, 20 Mar 2024 18:32:56 UTC No. 978274
>>978272
'intellectuals' isn't just people who bury themselves in books anon, broadly it just refers to people who habitually
inform themselves and think deeply about things outside their own areas of expertise.
You doubt moot wasn't intellectual? Ever heard him speak? Dude's pretty deep. Not every 4chan'er will be some /b/tard edgelord.
Just head over to /sci/ and take inventory of what kind of minds burn time posting on this site, perhaps you've spent too much time on them containment boards.
Anonymous at Wed, 20 Mar 2024 20:45:27 UTC No. 978278
>>978274
You don't know what an intellectual is, you are wasting your life with every passing second pretending you are doing something useful here, goodbye.
Anonymous at Thu, 21 Mar 2024 01:33:18 UTC No. 978297
>>975094
>>975099
1) images and videos are way better than you seem to be aware of or acknowledge
2) an idiot can see that, just based on the rate of progress, it will be perfected soon and be able to make genuinely novel content
you're in denial
Anonymous at Thu, 21 Mar 2024 12:25:41 UTC No. 978317
>>978297
>it will be perfected soon and be able to make genuinely novel content
no, it wont because big corpo would never allow anything to go against its big corpo message
Anonymous at Thu, 21 Mar 2024 13:05:34 UTC No. 978318
>>978297
>it will be perfected soon
ONLY TWO MORE WEEKS TRUST ME BRO™
Anonymous at Thu, 21 Mar 2024 14:32:15 UTC No. 978321
>>975073
It's not a roll if the people making the "ai god" are retarded and greedy as fuck. garbage in garbage out, this shit is just a force multiplier for people who are already fucking everything up.
These people do what they want, and get the outcome they expect, which is making everyone's lives worse and an extra dollar in their pocket. Except now they can blame it on AI if people actually get mad enough about being continually fucked in the ass to do something about it (they won't).
None of our problems are from a lack of computing power, except shit like nuclear fusion which OpenAI is not going to be able to help with at all lmao.
>>978271
"it actually takes creativity and intellect to be a proompter"
Long paragraphs of cope isn't an indication of a high intellect.
Anonymous at Thu, 21 Mar 2024 16:09:29 UTC No. 978326
>>978321
>Long paragraphs of cope isn't an indication of a high intellect.
To state something substantial that reveal valuable information on any given topic usually takes some space anon.
If what you have to say takes less than a paragraph to formulate, chances increase it's just junk/noise type data.
What you get from current day LLM's will very much be a reflection of your input, a person with a poorly nuanced or limited vocabulary
won't get as much back from a system like chatGPT as an adept language user because what it does is a synergy between the input and the output.
The systems of tomorrow may grow smart enough to answer the spirit of the question in language accessible to the level of the prompter
but as they exist today they very much latch on and extend the ideas you feed it.
I too cringe at the idea of someone considering themselves a 'prompt artist' while using generative AI to synthesize images, but there's skill involved.
Not anyone will be able to make these systems dance and sing.
You look at the prompts people are using to generate the highest quality specific output and they're downright autistic.
Anonymous at Thu, 21 Mar 2024 16:28:13 UTC No. 978331
>>978326
>Not anyone will be able to make these systems dance and sing.
If you have a task specific enough to require that much input, you are going to be smart enough to figure out the prompts or else you wouldn't be doing the task to begin with.
And in any case, whatever intellect is required to use these things, it's going to be less intellect than you would need to just write/make the thing yourself without the AI.
So bragging about skill or creativity for prompting is fucking embarrassing, particularly to those of us who can do these things ourselves.
Anonymous at Thu, 21 Mar 2024 17:10:49 UTC No. 978335
>>978331
>it's going to be less intellect than you would need to just write/make the thing yourself without the AI.
Absolutely, but even in delegating work you need to know what you want, and in the dark future of generative AI that can build entire worlds on the cheap
the mind of an architect who understands a lot about how he wants all his pieces to fit will be the kind of mind that will rise to the top
in prompting the machine to do his bidding.
>So bragging about skill or creativity for prompting is fucking embarrassing, particularly to those of us who can do these things ourselves.
Agreed. I just don't think the people who will create/consume AI art will care about how we find them embarrassing.
They will look at us like we looked at people who told us to learn how to count without the calculator, like we're asking them to waste
their time learning something difficult slow and ineffective that is more error prone and no longer matters in practice.
Anonymous at Thu, 21 Mar 2024 17:16:08 UTC No. 978336
>>978335
>build entire worlds
>he thinks what is on a ram stick is an entire world
point and laugh at this idiot
Anonymous at Thu, 21 Mar 2024 17:21:25 UTC No. 978338
>>978336
You didn't see what Sora did? The amount of depth and spatial coherence that is present there suggest 3D worlds won't be far off from being synthesized now.
Esp given how much depth data is present in the diffusion models when people feed those images into algorithms that reconstruct 3D space from images.
They already understand depth and parallax to a scary degree. If you think I'm an idiot in stating this it's because you have no fucking clue what already exist today.
Anonymous at Thu, 21 Mar 2024 17:22:50 UTC No. 978339
>>978338
>that world word again
Anonymous at Thu, 21 Mar 2024 17:26:39 UTC No. 978340
>>978339
>?
You don't know what a 'world' is?
Anonymous at Thu, 21 Mar 2024 17:27:12 UTC No. 978341
>>978340
huh?
Anonymous at Thu, 21 Mar 2024 17:39:01 UTC No. 978342
>>978341
https://en.wikipedia.org/wiki/Virtu
You prob seen some, they're pretty common in various entertainment media you probably consumed if you've spent any time on this planet.
Anonymous at Thu, 21 Mar 2024 17:42:23 UTC No. 978343
>>978342
Go do something productive with your life, second life kid
Anonymous at Thu, 21 Mar 2024 18:27:52 UTC No. 978344
>>978335
>the mind of an architect who understands a lot about how he wants all his pieces to fit will be the kind of mind that will rise to the top
There is simply not going to be a Randian superman prompter. Aside from the fact that prompting only appeals to uncreative dullards, it doesn't make any sense in our world and our economy with this technology. The people who are at the top of any creative industry are the people who can make personal connections, with their peers, with bosses, with audiences. That's not going to change, and if prompters want to do that they're already shooting themselves in the foot by being a prompter. if someone miraculously makes something good with AI, people will say "wow this technology is great", and the prompter will (rightly) get no recognition.
And if you don't want to connect with an audience, the only ai jobs are going to be minimum wage at best. More likely it's going to be people in the third world who can live on pennies a day doing the prompting.
>Agreed. I just don't think the people who will create/consume AI art will care about how we find them embarrassing.
lol they fucking do, look at Shad Brooks he got so ass-blasted about not being respected as an artist that now he's blaming all his problems on Woke.
This shit just straight up sucks, there's no upside for anyone here but the likes of Disney, and the handful of grifters willing to shill ai art on etsy.
Anonymous at Thu, 21 Mar 2024 18:43:28 UTC No. 978345
>>978344
>there's no upside for anyone here but the likes of Disney
Disney is in the process of going bankrupt
Anonymous at Thu, 21 Mar 2024 20:07:00 UTC No. 978348
>>978344
>lol they fucking do
Only because it still kinda sucks in all but the most exceptional cases, we're at the 'cool techdemo' stage atm but actually implementing it for anything is premature.
However that's just not gonna stay the case, I've seen plenty examples of AI art that I can't deny that is better than most of things I've seen by real artist.
It's quite obv to me that is gonna be the norm of tomorrow and what's beyond that is something that will eclipse all of us.
I'm an artist and I myself ultimately don't care if art I consume is made by a human or a machine, the value in making art myself is the intrinsic joy of creation.
I don't have access to the internal process of my favorite artist any more than I have access to what the machine spits out, ultimately I only care if I find it good.
If I think that way, why would any consumer of art care how the content they enjoyed came about? They won't.
Soon as art AI delivers is consistent and more appealing than human made one they'll switch to the cheap and abundant option.
Anonymous at Thu, 21 Mar 2024 20:26:17 UTC No. 978351
>>978343
>Go do something productive with your life, second life kid
What does that even mean? You think I'm some kind of reincarnated being like the Dalai Lama?
If I have the accumulated wisdom of multiple lifetimes why would it matter if I'm a kid or not being productive. I should have all the time in the world.
Anonymous at Thu, 21 Mar 2024 20:27:56 UTC No. 978352
>>978348
>I'm an artist
prove it
Anonymous at Thu, 21 Mar 2024 20:49:48 UTC No. 978353
*crickets*
get the fuck out of here kid, you dont know shit about making art
Anonymous at Fri, 22 Mar 2024 06:13:29 UTC No. 978376
>>978352
>>978353
Posting current work would dox myself but how about this, this is my main work drive and that's the backup of max-files from my 2016 computer tallying 89gb of files. As you can see I've been at this a while.
Anonymous at Fri, 22 Mar 2024 08:45:31 UTC No. 978383
>>978376
The only thing you've proven to me so far is that you are an ngmi with a very high opinion of himself, post your actual work, I don't give a shit about your folders
Anonymous at Fri, 22 Mar 2024 09:38:27 UTC No. 978385
>>978383
Well you've now seen proof how I've invested my life in this and that art is what I do. I'm fully exposed to all the ill effects of generative AI.
Whether you wanna believe I'm a good artist or a bad artist doesn't change how I am an artist.
You've been provided proof I speak from the vantage point of someone who's interests are very vested.
If you wanna keep moving the goal-posts adding additional qualifiers to who is entitled to have opinions that's on you.
Just realize I'm not saying what I'm saying about AI because I wish for any of this to be true.
I'm saying what I'm saying because it's what I deem to be true based on careful examination and observation of the on-goings.
Anonymous at Fri, 22 Mar 2024 10:00:10 UTC No. 978386
>>978385
>Well you've now seen proof how I've invested my life in this and that art is what I do
I have not
Anonymous at Fri, 22 Mar 2024 10:08:57 UTC No. 978388
>>978386
Well either you have or you believe I sit here and fake folder structures and blank out personal information to develop some fake persona in a random 4chan thread.
I'll leave it as an exercise for the reader to Occam's shave that one for themselves.
Anonymous at Fri, 22 Mar 2024 10:40:33 UTC No. 978391
>>978388
You're pathetic. You suck fucking ass, cris
Anonymous at Fri, 22 Mar 2024 11:05:39 UTC No. 978393
>>978391
>You're pathetic. You suck fucking ass, cris
shut your mouth, I'm pretty glorious..
plus I've never sucked ass cept maybe the occasional one of a girlfriend but that's just something that happens in the heat of the moment and doesn't count.
Also I'm clearly not 'cris', you can easily tell how I'm not cris because you can't insult cris at all without having him explode in some massive multi-post hissy-fit
where he's insulting anything and everything all at once.
Now, do you have anything of substance to add anon or are you just getting off at inventing your own head-canon while insulting people?
Anonymous at Fri, 22 Mar 2024 11:10:53 UTC No. 978394
>>978393
Oh but if you mean 'You suck fucking ass' that could actually be true, I'm not into anal at all and have actually only done it once.
I've always said that buttsex is like urbex; it's a lot cooler on film. In person it's like moldy damp and smells weird.
If that's how you intended it I must yield you might be onto something with that critique.
Anonymous at Fri, 22 Mar 2024 11:14:47 UTC No. 978395
Anonymous at Fri, 22 Mar 2024 11:40:10 UTC No. 978396
>>978393
>explode in some massive multi-post hissy-fit
That's what you've been doing for the past 50 posts, my personal lolcow
Anonymous at Fri, 22 Mar 2024 11:55:44 UTC No. 978399
Things are not going well for cris and cris-kind.
Perhaps we should require all "people" talking about AI on /3/ to submit verified Work before engaging them further lest all over our time gets collectively wasted
Anonymous at Fri, 22 Mar 2024 12:38:15 UTC No. 978406
>>978396
Posting is not the same as throwing a 'hissy-fit' anon, I've been nothing by cordial to my detractors despite volumes of name-calling.
Take you for example, you call me 'your personal lolcow' looking to get a rise out of me but the response you received here was cool calm and collected.
If you find it's 'lolcow' levels of funny just how someone authors text to you isn't this danger-close to how babies are amused when someone go "biggabooboo!" infront of them?
My psychology is AR500 levels bulletproof anon, shit thrown into my face simply slips off my shoulders like drops of water over the hydrophobic coat of a penguin.
Anonymous at Fri, 22 Mar 2024 12:41:10 UTC No. 978407
>>978406
>My psychology is AR500 levels bulletproof anon
see
>wont post work
>time wasting irrelevant posts like
>>978393
>>978394
good morning, cris
Anonymous at Fri, 22 Mar 2024 12:51:25 UTC No. 978409
>>978407
Not waste of time, those posts made me lol.
>I'm pretty glorious
>My psychology is AR500 levels bulletproof anon, shit thrown into my face simply slips off my shoulders like drops of water over the hydrophobic coat of a penguin.
NGL that is pretty glorious, we need more people like that.
Anonymous at Fri, 22 Mar 2024 13:24:45 UTC No. 978413
>>978406
Post something again my little lolcow
Anonymous at Sat, 23 Mar 2024 02:13:22 UTC No. 978472
>>975140
Until ai says the n word instead of nukes the planet it will never be smart.
Vc j00
Anonymous at Sat, 23 Mar 2024 02:18:08 UTC No. 978473
>>975278
>Lobbying for less regulations
More. They're lobbying for more you completely cucked kike
Anonymous at Sat, 23 Mar 2024 02:20:21 UTC No. 978474
>>977913
You are as dumb as the 8 year old holding that doll and if you thought Lisa was supposed to be the voice of reason in the show you basically weren't watching any of the original 8 seasons of the Simpsons.
Anonymous at Sat, 23 Mar 2024 02:25:33 UTC No. 978476
>>978158
>Chat has no goals
It has plenty of goals. For one it wants to make a video game about time traveling.
Anonymous at Mon, 8 Apr 2024 12:12:47 UTC No. 979955
the fuck happened in this thread