Calling people Luddites is the new “are you triggered yet, are you triggered yet” Yeah man, I’m so unbelievably afraid it’s crazy. That predictive text has me shaken to my core.
Chris Marker, whose entire career was about exploring the mysterious connection between image and memory, says "Memory is not the opposite of forgetting, but rather its lining." A data set of information is not the same as memory.
There is this idea called The Empty Internet. That even before AI proliferation, very, very large amounts of Internet activity were just bots. Could be as low as like 30-35 percent. But the high estimates are quite high indeed: 80 to 85 percent of every tweet, comment, transaction, product review, like, channel subscription, etc comes from fake activity. That humans are actually a real rarity to encounter online.
I’m developing a horror movie atm, and I used midjpurney to make my deck. It was tough and if I wanted an original idea visualized almost impossible. It doesn’t “think” like a person, it doesn’t know film language, or understand basic direction. Writing prompts is a pain in the arse. Lots of trial and error. It’s like a game of telephone with an alien.
Just a few years ago, AI image generators were only doing style transfers, e.g. they could take a pattern from one picture and put it on another. Before then, AI art consisted of basic fractals and screensaver patterns. AI isn't going to 'replace Hollywood' in 2025, but by the end of the century, it's definitely going to be capable of impressive and frightening things. This doesn't make it good or bad, or a form of creativity, but it also doesn't mean AI art can't have some intrinsic or aesthetic value attached to it.
Exactly my thoughts. The rate of improvement is astounding. AI video looked like nightmare fuel less than a year ago. Now it can convincingly be used in short ads without the average person even noticing anything strange.
I personally like how Twitter has nazi types talking about how blacks should be “grateful” and Andrew Tate muslim types preaching how to “treat ya woman” it’s good stuff all around!
The issue with AI comes down to who's developing it and for what purpose. The problem we collectively face is capitalist AI and ultimately capitalism in its entirety. We exist during the highest stage of capitalism the age of imperialism and these different imperialist powers are developing AI for their own capitalist class interests and not for the class interests of workers. AI for all intents and purposes is a tool. A very complex and powerful one at that however it's still a tool nonetheless. Moralizing the technological field of AI as a whole one way or the other (many even in just this comment section are reactionarily opposed to AI) is pretty ridiculous since it can be used in both beneficial and harmful ways. It's inherently relative and neither good or bad.
I hate to break it to you, but Vaush also has tons of loli that was drawn by hand. As it turns out, the computer can't actually do anything on its own. It needs a human to tell it what to do, and can only generate art by looking at images that pre-date the existence of AI.
In a totally random topic not related to garbage Ai images, anyone else remember Ben Shapiro's image of an allegedly burned baby? (that definitely didn't have text blurred out to cover up the garbled Cthulhu words that get generated by Ai software)
Why do people assume tech CEOs create the tech they sell? Nobody assumes Henry Ford personally designed the Model T...right? Is it because Musk and Zuck are such dummies in public that we assume they must be bringing something to the table?
Yes, people do believe he that. Remember all those articles describing Musk as the real life Iron Man? Remember when he had a fucking cameo in one of those movies, where he and Tony Stark complimented each other for being genius's? The revolutionary inventor turned Titan of industry is a classic American archetype that we've been fed for generations by Hollywood. Of course, they've always been just thieves and Liars, but it's these successful thieves and liars that founded and owned Newspapers and Film studios. It's capitalist propaganda. The tech ceo is just the latest iteration of the myth.
Because the entire point of a CEO is to take credit for the work of their laborers. You could take all their money, put it back into the company, get rid of them, and *NOTHING* of value would be lost.
Of course tgey think Elon is out there snapping Tesla’s together like Lego sets, sleeping on the factory floor, designing rockets in his dreams and receiving direct messages from god. Obviously I’m joking but I can 100% guarantee there’s pretty if ppl that think this
The price of the Vision Pro isn't at all surprising when you remember that they sell a monitor stand for $1000 Not a multi-monitor stand with an auto-turret to kill unauthorized users, just one stand for one monitor.
I think doubting future capabilities of technology is just setting yourself up to be wrong. The more interesting point is why are we even utilizing technology this way. For what? Just to churn out more inane hemorrhoids ads about progress. For more dreck for little iPad babies to be mentally maimed by? It’s a fantasy to consider technological progress as any kind of progress that benefit us as humans anymore. We’ve been accelerating a trajectory to outmode ourselves for some time now. The question of to what actual end that is good for us has been lost in the ubiquity of financial capital. Something something dialectics something something Marx was more right than he could have ever known. *sniff
The problem with that though is that the so called 'ai' we see in common use nowadays cannot and is not what it is marketed as. It cannot think, it cannot know, and it cannot be creative. It certainly has its uses and can do a lot of very important and interesting stuff but it is not in any way the intelligence it is named after. Perhaps one day we will have such technology but if we do, it will have to be made in an entirely different way than the language learning models we have now
@@Robot_Eva confusing AGI with GenAI is definitely a problem that folks within the AI education community have been working at. Most of the education and training material I’ve reviewed is pretty clear about the difference, but that doesn’t stop the public’s general lack of understanding from being exploited unfortunately. Education/awareness is usually the answer to scams and fraudulent marketing though.
"It cannot think, it cannot know, and it cannot be creative." -- unless you define those words, nobody has any idea what you're saying. That's been the whole problem with thinking about AI from the start, and philosophy long before that. @@Robot_Eva
@@Robot_Eva what you are thinking about is "general intelligence". I get your point and the Chapo crew's that machine learning, or however people want to split hairs and call it, is still in this nascent state where it can at least be mocked still. And that all these people tech doomer posting are being silly and sophomoric in their takes. When I address this kind of posturing though people tend to miss my point. And it is literally the postmodern cliche notion of "people no longer being the ends of things and things in and of themselves take on their own ends". It's been happening for quite some time now like I've said. But all this generative language model stuff is another acceleration towards this notion. Clearly seen in how it is brought out in the anxiety of people reacting to this promised future of being outmoded. With it's proponents also either just coping or actually too far deep in the Koolaid of neoliberalism to see this dash towards a further erasure of what it means to be human under free market capitalism. I don't care if in the future ads will all be generated by large language models. The problem from the get-go is that there are even jobs at all for spinning up some fantasy to sell for pills that make your dick hard. "Maybe if your dick was harder you would be a better father". Synthesize that idea into a prompt and save 5 million dollars in production costs. It's ridiculous. And it's just going to get even more nonsensical. The bloat of tech and the market in general should only go so far before people start questioning the "value" of it all.
Image generators and the like were around a long, long time before they were introduced to a mass audience. Been improving and will keep improving, dunno why there's such a doubt they can become indistinguishable on some level. Lotta research happening on machine learning a lot deeper than image generators
It's ridiculous to say that generative AI for images and video hasn't gotten any better in the last year. It is obviously, extremely better at producing coherent and convincing images than at this time last year. That doesn't mean that it's a perfected technology (or that this is indeed useful technology at all), but at least be honest about the pace of development here.
AI has gotten superficially better in the past year. Chapo is going to feel so silly in another year when the lady who is part bed and part cat looks less like a product of ketamine.
A.I. right now is I'm it's infancy and early childhood. The problem are these psychotic tech bros are treating it wrong like an abusive parent. We all know what happens to children in their teens and early adulthood after living in an abusive household.
@@Fucyallfr some critical thinking would be nice. We get Bevis and Butthead style commentary mostly calling Elon Musk names with no thought about society and the possibility of changing it, the rationality of events in the context of capitalism, etc. Also they are constantly saying things that are factually untrue.
This guest is not accurately representing the technical details of these models. At one point he claims that in generating text or images these modela are just refrencing a huge dataset. The dataset is not refrenced at runtime. OpenAI could delete all the data they took from the internet to train GPT4 and GPT4 will not be affected even a tiny amount. You can think of it as a giant stack of numbers that conditionally respond to some context. You pass a bunch of context with some deliverate ommisions past those numbers and check how well the numbers guess what pieces are missing. Then, you continuously nudge all the different numbers in the direction that causes them to make better guesses. The whole assertion that they cant think or that hallucinations are an integral part of their operation is extremely dubious, especially when spoken with that level of confidence, because we just don't know how the capabilities we see on the surface are actually achieved using that massive stack of numbers that was nudged into performing as well as these models currently do. To predict perfectly the data they are trained on, they would have to represent the causal processes which originally produced that data. Like, if in the dataset there's a record of a series of scientific demonstrations followed by the observed data, in order for the giant stack of numbers to guess what observations follow from the description of the procedures, it has to reason about the physical processes which caused those numbers to be written down in the "results" section. More concretely, if you have a description of mixing specific concentrations of acids and bases in a chemistry experiment, followed by a description of the resulting amounts of the produced chemicals, the AI has to be getting better at reasoning about chemistry abstractly in order to better guess those results. In this interview I see someone who has a take about AI that theyve arrived at because they don't like the attitudes, politics, and general vibes of the people who "cheer" AI on, and then went though a bunch of pop culture sinplifications of what these AI programs do, then took away a misinterpretation of that information which - if it was true - makes the people they dislike look like credulous idiots. It reminds me of those new age gurus who misunderstood quantum mechanics to mean that consciousness was some sort of immortal property of the universe, and that all minds were entangled together in a great web, and "quantum immortality," "universal consciousness," "minds entangled across all time," etc etc. They wanted a nice beautiful story about how people are really special and we'll all live forever as cosmic waves or something, and they read a bunch of pop science stories about quantum mechanics and string theory, but didn't bother to actually figure out what the people who wrote those papers meant when they said "the observer collapses the waveform" or "quantum immortality and the many world interpretation." Similarly, this guy seems to want a story about how tech people are totally wrong about how useful or promising this tech is, and how easily impressed they are by something that is obviously just a flimsy illusion. Then, went and heard some 5 minute UA-cam videos or read a few Time articles, and came away with a misinterpretation of what this technology even does at a basic level. He even said "average general intelligence" was the thing the field was seeking. Which just isn't what the acronym "AGI" stands for. Thats just not a mistake someone would make if they'd learned enough about this topic that they would be able to make bold categorical statements about how these things work and what they'll fail to achieve by what stage of development. The damn thing is, I agree that the people in charge of these labs are short sighted idiots who are doing all of this in such a proud, loud, and snobby way because someone called them a naive fool for thinking AI was important or possible back in 2012. There's a ton here thats completely valid to criticize. I'd point my finger at the fact that these people can't seem to agree that it might be important to be able to agree on where all of this is heading and actually plan in a coordinated and sensible fashion for how to deal with that. The problem is, not every criticism is going to inspire the exact right policy that we need to take to fix this situation. You can't just say "clearly these people suck in a bunch of ways, why shouldn't I just use whatever argument seems most damaging to them?" The way you pick the take that will actually cause people to think of beneficial improvements to the situation is to try and succeed at making whatever you say actually true. I know its hard to actually figure out what's true about this technology, because that involves listening to people who are very smug and talk using a bunch of terms that aren't actually common knowledge. But, y'know, kinda notice how a bunch of math weenies are making huge waves in the world, and maybe be a little curious what these weirdos are actually talking about. I guess this just isn't the crowd that you're supposed to say something like that around? It feels like Matt, Amber, and Virgil are more of the type to go "okay, no, what this person actually means is this, and if you look at it from this angle you can kinda see why that would be true, but they seem to be missing X and Y and Z." In other words, I predict if I got to whisper in the ear of Chapo Trap House in general and pointed the show at a website like LessWrong or more specifically The Sequences, like for a segement like a reading series, there woukd be a quick scan of some material, then they'd conclude that it's not an easy target to mock nor particularly entertaining to try and comment on. There'd be some sort of reflex that disinclines this group from engaging with material which isn't just so full of nonsense that it's easy to talk disparagingly about it for an hour without running out of things to mock - and not feeling bad about mocking someone for having something like a bad hairstyle, because they're so awful on other topics that it feels like there's a huge budget for spreading that negativity around to other less independently valid areas of criticism. Am I wrong? Will Chapo turn out to have the correct take here? Does it seem kinda like, on a gut level, that they'll actually miss some trend going on here that'll become important in maybe even less than a year? Do you feel that hint of doubt at the back of your head? Did you sense that quiet note of confusion that sorta flitted around behind your eyes when some of these statements were being made? That a huge fucking alarm bell in my book, it means you can kinda tell this isn't even a totally solid slice of the real story. It's the sensation of seeing a mirage. You want that to be water on the horizon, and you may start frantically lurching towards it, overwhelmed to explosive effort with optimistic certainty that you're interpreting things correctly - but, if you were more aware of what was going on in yohr head when you first widened your eyes at the shimmering sight in the distance, youd have noticed that you didnt even fully believe the story you told youraelf about thay definitely being water. In the back of your head you knew what seeing water felt like, you could tell it wasn't a certainty when you first spotted the shimmer, you may have even had a sinking feeling in your stomach when you started moving towards it. That's what I think many people will feel when they listen to this. Maybe it's even what Chapo is feeling when they say this stuff. It's a hard sensation to identify and react to, but it's one of those you ignore at your own dire peril.
Why does it matter what they think? They arent computer scientists or engineers, and as you pointed out - some of the basic misunderstandings from the guest indicate he isn’t very well informed on AI/ML either. I just don’t think that their observations about the over hype of generative AI will lead to “peril” or however your conclusion ended up
How are they so definitive that this technology isnt going to radically improve in a decade or 2? When I was young we had a lot of these new techs that sucked at launch (motion sensor duck hunt etc.) now theyre so good the biggest selling console was based on motion sensors. VR looked like it sucked for a while but again we can see the Apple headset that by reviews at least suggests theres something here now. I wouldnt remotely be so confident in dismissing tech because were in the duck hunt era.
Are we still in the "Duck Hunt era" of the Hyperloop then? Is that just guaranteed to even *exist* in 10 years, let alone outpace existing public transportation methods, simply because that's *your* understanding of how technology develops? Listen to the episode again. They acknowledge the argument that AI is "improving." They are unimpressed. That the Virtual Boy may have been technically superior to Pong does not mean it's an "improvement" over the latter.
@@kingofsting19 No because the hyperloop (elons crap right?) is nonsense, theres nothing innovative there. They dont remotely acknoledge it beyond a lay up question what do you say to ai improving etc.... which this guy responds with the most insanely confident statements that this stuff is going nowhere. Its geniunely insane to argue today that you are certain that this technology is always going to be dogshit. Theres an absurd amount of capital flowing into this globally, someone somewhere in over our life is probably going to get this technology to a place that is geniunely impressive to a mass consumer audience.
@@MichaelHachey Yea but ignore those weirdo bluecheck freaks, theyre morons moving off nfts to some new grift, but my take from this guys analysis is basically ai will never replicate something that creators do today in the creative space because of all these flaws. I could have happily argued that in the 90s when I played duck hunt that motion sensors will never get anywhere but they did 20+ years after.
@@Mezzo396 The US military spent $2 billion trying to build the B-2 bomber and it still can't fly in the rain or eject without killing the pilot at least a quarter of the time. It's not "geniuenly" insane to not subscribe to the magical thinking that enough time and money *"must"* correlate to a certain level of technological advancement. Sometimes all you do is push money around and find out why something doesn't work, like the Hyperloop. Also, who cares what a "mass consumer audience" thinks, bootlick? A mass consumer audience once decided "pet rocks" were a hot item. Come back to me when these AI do something that justifies the amount of money being spent on them, not just be a new thing for the billionaires to speculatively invest in.
This guys sounds like he just got done memorizing "The Power of Positive Thinking." Everything he says is just so uplifting and inspirational! This is what the world needs right now! More positivity! That's the spirit!
As someone who works in AI, the first minutes seemed strange, because one thing AI really brings out for us is that creativity is not a skill. That whatever AI creates has the indelible stains of the mass of culture that produced it. It never seems "fresh" somehow, not without a lot of creative human intervention. That said, I think everyone is underestimating what could and probably will happen. I'm not a booster of AI. In fact, I hate dealing with it, and I hate computers, too. I'd love nothing more than to see expectations collapse as we hit some fundamental limitation, but they won't. Shit's for real. In a lot of ways, it already really does things the way we do them. There's no fundamental difference between the neural networks in our head, and those in AI software. The only big limitation now is that the AIs have few ways of interacting (and therefore learning about) with the real world, thank god, because robotics has (comparatively) lagged behind and remained expensive. So it only knows the abstract world that's on the internet. It can write a rap about the sequence of operations needed to make a radio circuit with any random collection of components, but it wouldn't know how to even "cut" a wire. But if it was in a robot with good abilities, it could literally figure it out just by looking at videos on the internet.
@@obotopo who hurt you lmfao? Dude doesn’t deserve to be called a vampire just because you’re on a soapbox. Do you really think that anyone with a tech job is incapable of creativity lmfao?
@@ultravioletiris6241Yes. Lmao. “Tech” and technology are not analogous anymore. Most tech bros hired somebody else to do the creative work that they sell as if they made the advancements themselves. This is why nobody takes the “tech” sector seriously, if you want real creative endeavors in the 21st century, go to local businesses and companies where there are actual stakes or academia.
I REALLY dislike pretty much any form of generative AI. But "it looks bad" is a terrible argument. This tech has obviously made leaps and bounds over the past two years. Frankly, I think you're kidding yourselves about the video generators like SORA. That shit will absolutely fool the average moron on Facebook. I give it a month tops until it's used to generate some kind of horrifying atrocity propaganda.
Nah. It all looks smooth and dead. No soul. You could maybe argue it looks kinda cool aesthetically but it's all just so empty looking. For sure it will be used for evil though, absolutely.
If AI ever does get more advanced and is in android bodies with self awareness I hope there is a civil rights movement for AI autonomy. Yes I read too much scifi, watch Chobits, and play Detroit Become Human.
I can haz cheezburger?? EPIC FAIL! Ok, this wins the Internet ☝️ ROFLCOPTER does the bacon narwhal? this post gave me cancer EPIC WIN edit: thanks for the gold kind stranger!
@@Praisethesunson It’s not about being a simp. If it’s important to not fall for the AI hype, it’s equally important not to spread misconceptions the other direction too.
Yeah man not adopting shitty broken technology is just like that time a bunch of workers went into factories and broke the machines. You're so smart. *Yuh man just have faith it's gonna get so good. Technology gets better so the Juicero is gonna be the juicer of the future maeen.*
I think most people who understand the implications are pretty concerned about this. There's no good that will come from from taking the human element out of creativity.
@@biscuithammer00 Anyone who cries about "taking the hu.an process out of creativity" doesn't understand how AI art or actual creativity works. AI doesn't just make itself, numbnutz. You need a creative human to tell it what to do. AI art cannot exist until I use my human creativity to type a prompt into the text box.
@@abreathingcoffin8089 Yes. Being afraid of the computer because it might take your comissions for drawing furry pr0n is, in fact, exactly like being afraid of cars because they might take your job as a hore drawn carriage driver.
They're gonna eat their words on this one in a few years. AI is going to become much more capable and prevailing in our lives, probably in a negative way. These guys should really stick to reviewing Kubrick films because they clearly don't know shit about the field of AI.
@@abreathingcoffin8089 I'm currently getting a masters degree in electrical engineering and I use machine learning techniques for radio signal analysis. I've also been following the field for years and talk to friends that work directly in AI. That being said, I don't think you need a masters degree to see that the progression from early ML models, GPT-2, GPT-3, chat GPT, and now Sora is continually trending towards something we would consider intelligent, general, and even creative.
@@mattcorrigan6639but you sound like you know things your talking about… if anyone here had the knowledge to confirm or understand what you said then we could determine if you actually know what your talking about or just know sone words that fool us…
Oh wow, is the country built on forced military service, propped up by the largest weapons-producing country in the world, winning its illegal war against its own minority population, who aren't allowed to leave their fence prison or even fish in their own ocean, let alone have a military? Epic, bro. You definitely *don't* look like a massive bitch for being excited about that.
Your stupid algorithm can't think or be creative, you're just deluding yourself with regurgitations of shit that already existed and was made by people
Hey look, some teenage edgelord who thinks his disdain for his fellow man makes him a freethinking genius and not a boring cliche just looking for an excuse to not contribute to the world.
It is correct. There are many different ways to say that expression, and this one is fine. Besides, native speakers know the difference between spoken word and written word, and since English is notoriously complicated, much leeway is given to native speakers.
Literally proper grammar. Will speaks with a half NY/half mid-Atlantic accent and you don’t think that half his time in primary school was just getting wrapped on the knuckles for improper grammar? Come one jack
what frustrates me is these AI video tools obviously cannot do any of the pointless hollywood fanservice stuff these people say they will, but they could absolutely be used to make some incredible experimental stuff, we're just not seeing a whole lot of that yet. one incredible example of the potential is this michael jackson film, beautiful haunting work of art that uses AI to great effect, exactly the kind of thing we should be seeing more of: ua-cam.com/video/P2zPO-MFBkQ/v-deo.htmlsi=PlPNPfEty7f29eQ1
Yup the marketing and hype is definitely over the top. Tech does this a lot for various reasons such as to seek investments or customers, and now you get to throw in even more transhumanist mysticism than usual.
These AI people think art is a product not a process, because they want to sell it not enjoy it.
Elon Musk is exactly what a semi-self-aware AI would be like
Calling people Luddites is the new “are you triggered yet, are you triggered yet”
Yeah man, I’m so unbelievably afraid it’s crazy. That predictive text has me shaken to my core.
lol at Chapo attracting Musk enthusiasts to post on this video.
Chris Marker, whose entire career was about exploring the mysterious connection between image and memory, says "Memory is not the opposite of forgetting, but rather its lining." A data set of information is not the same as memory.
guys I made an AI version of this episode of what it would sound like it matt was back.
Can u make it if Matt was black? And me look chinese while ur at it
The media not pointing out that Musk is an idiot is such an emperor has no clothes example that it is kinda cliche
Given that the internet is a series of tubes we've technically had AI since that old Windows screensaver that generates a bunch of pipes
I am genuinely terrified of large chunks of the internet becoming completely unusable due to the proliferation of ai schlop.
erect the blackwall
There is this idea called The Empty Internet. That even before AI proliferation, very, very large amounts of Internet activity were just bots. Could be as low as like 30-35 percent. But the high estimates are quite high indeed: 80 to 85 percent of every tweet, comment, transaction, product review, like, channel subscription, etc comes from fake activity. That humans are actually a real rarity to encounter online.
I’m developing a horror movie atm, and I used midjpurney to make my deck. It was tough and if I wanted an original idea visualized almost impossible. It doesn’t “think” like a person, it doesn’t know film language, or understand basic direction. Writing prompts is a pain in the arse. Lots of trial and error. It’s like a game of telephone with an alien.
Just a few years ago, AI image generators were only doing style transfers, e.g. they could take a pattern from one picture and put it on another. Before then, AI art consisted of basic fractals and screensaver patterns.
AI isn't going to 'replace Hollywood' in 2025, but by the end of the century, it's definitely going to be capable of impressive and frightening things. This doesn't make it good or bad, or a form of creativity, but it also doesn't mean AI art can't have some intrinsic or aesthetic value attached to it.
Exactly my thoughts. The rate of improvement is astounding. AI video looked like nightmare fuel less than a year ago. Now it can convincingly be used in short ads without the average person even noticing anything strange.
I personally like how Twitter has nazi types talking about how blacks should be “grateful” and Andrew Tate muslim types preaching how to “treat ya woman” it’s good stuff all around!
The issue with AI comes down to who's developing it and for what purpose.
The problem we collectively face is capitalist AI and ultimately capitalism in its entirety. We exist during the highest stage of capitalism the age of imperialism and these different imperialist powers are developing AI for their own capitalist class interests and not for the class interests of workers.
AI for all intents and purposes is a tool. A very complex and powerful one at that however it's still a tool nonetheless.
Moralizing the technological field of AI as a whole one way or the other (many even in just this comment section are reactionarily opposed to AI) is pretty ridiculous since it can be used in both beneficial and harmful ways. It's inherently relative and neither good or bad.
Completely agree, great to see someone approach this like a marxist.
13:35 Make it a Slavoj Zizek AI. Whenever you ask it to do anything it just replies with: "I'd prefer not to."
Me: "I want a sexy woman in a bikini."
ZizekAI: "What is a woman?"
Me: * sighs *
that'd be a Herman Melville AI
I cant wait for Matt to come back
Vaush had AI generated loli…. Al generated loli really, really black pilled me. Let’s just end this god forsaken experiment
Still on Vaush huh?
Rent free
@@just_a_turtle_chad ayy it's turtle chad
I hate to break it to you, but Vaush also has tons of loli that was drawn by hand. As it turns out, the computer can't actually do anything on its own. It needs a human to tell it what to do, and can only generate art by looking at images that pre-date the existence of AI.
@@just_a_turtle_chadBro talking about rent free when he comments on every single chapo video
What experiment? The internet?
God, felix is so funny
Wait Ed Zitron's elon musk is actually pretty good
disregarded this guys opinions on everything when he said that muses rendition of house of the rising sun was amazing
In a totally random topic not related to garbage Ai images, anyone else remember Ben Shapiro's image of an allegedly burned baby? (that definitely didn't have text blurred out to cover up the garbled Cthulhu words that get generated by Ai software)
Im sure there are some benefits to using ai for creative work but if i had to get rid of that to kill ai it would be a worthy sacrifice
Just wait till , James Cameron tries to put the AR thing in theaters .
Isn't "cut-scenes from 'Heavy Rain'" redundant?
any hack can type prompts, but there’s a big difference between being “creative” and being merely “generative”
Why do people assume tech CEOs create the tech they sell? Nobody assumes Henry Ford personally designed the Model T...right?
Is it because Musk and Zuck are such dummies in public that we assume they must be bringing something to the table?
Yes, people do believe he that. Remember all those articles describing Musk as the real life Iron Man? Remember when he had a fucking cameo in one of those movies, where he and Tony Stark complimented each other for being genius's?
The revolutionary inventor turned Titan of industry is a classic American archetype that we've been fed for generations by Hollywood. Of course, they've always been just thieves and Liars, but it's these successful thieves and liars that founded and owned Newspapers and Film studios. It's capitalist propaganda. The tech ceo is just the latest iteration of the myth.
Because the entire point of a CEO is to take credit for the work of their laborers. You could take all their money, put it back into the company, get rid of them, and *NOTHING* of value would be lost.
Of course tgey think Elon is out there snapping Tesla’s together like Lego sets, sleeping on the factory floor, designing rockets in his dreams and receiving direct messages from god.
Obviously I’m joking but I can 100% guarantee there’s pretty if ppl that think this
AI will NEVER write about how sixes hang in the door
What if this whole episode was made by ai...
first 20 minutes might as well have been
The price of the Vision Pro isn't at all surprising when you remember that they sell a monitor stand for $1000
Not a multi-monitor stand with an auto-turret to kill unauthorized users, just one stand for one monitor.
30:35 Another World mentioned
Love getting my twitter/Elon fix. Need more.
I think doubting future capabilities of technology is just setting yourself up to be wrong. The more interesting point is why are we even utilizing technology this way. For what? Just to churn out more inane hemorrhoids ads about progress. For more dreck for little iPad babies to be mentally maimed by? It’s a fantasy to consider technological progress as any kind of progress that benefit us as humans anymore. We’ve been accelerating a trajectory to outmode ourselves for some time now. The question of to what actual end that is good for us has been lost in the ubiquity of financial capital. Something something dialectics something something Marx was more right than he could have ever known. *sniff
The problem with that though is that the so called 'ai' we see in common use nowadays cannot and is not what it is marketed as. It cannot think, it cannot know, and it cannot be creative. It certainly has its uses and can do a lot of very important and interesting stuff but it is not in any way the intelligence it is named after. Perhaps one day we will have such technology but if we do, it will have to be made in an entirely different way than the language learning models we have now
@@Robot_Eva confusing AGI with GenAI is definitely a problem that folks within the AI education community have been working at. Most of the education and training material I’ve reviewed is pretty clear about the difference, but that doesn’t stop the public’s general lack of understanding from being exploited unfortunately. Education/awareness is usually the answer to scams and fraudulent marketing though.
"It cannot think, it cannot know, and it cannot be creative." -- unless you define those words, nobody has any idea what you're saying. That's been the whole problem with thinking about AI from the start, and philosophy long before that. @@Robot_Eva
@@Robot_Eva what you are thinking about is "general intelligence". I get your point and the Chapo crew's that machine learning, or however people want to split hairs and call it, is still in this nascent state where it can at least be mocked still. And that all these people tech doomer posting are being silly and sophomoric in their takes.
When I address this kind of posturing though people tend to miss my point. And it is literally the postmodern cliche notion of "people no longer being the ends of things and things in and of themselves take on their own ends". It's been happening for quite some time now like I've said. But all this generative language model stuff is another acceleration towards this notion. Clearly seen in how it is brought out in the anxiety of people reacting to this promised future of being outmoded. With it's proponents also either just coping or actually too far deep in the Koolaid of neoliberalism to see this dash towards a further erasure of what it means to be human under free market capitalism.
I don't care if in the future ads will all be generated by large language models. The problem from the get-go is that there are even jobs at all for spinning up some fantasy to sell for pills that make your dick hard. "Maybe if your dick was harder you would be a better father". Synthesize that idea into a prompt and save 5 million dollars in production costs. It's ridiculous. And it's just going to get even more nonsensical. The bloat of tech and the market in general should only go so far before people start questioning the "value" of it all.
@@bdinh3130👆guy actually gets it
A.I. ...? Is that for The Art Institute? I went to PDxAi (The Art Institute of Portland). I am now unemployed and live with my parents.
hell yea dude
Grok is woke
I’m gay
Im on my lunch break
Congrats.
The first gay AI chatbot , congratulations
Hot
nicee
'How good the Internet is now' 17:12
Gotta get right like Ed Zitron
Keep the intro music in the youtube uploads please.
Image generators and the like were around a long, long time before they were introduced to a mass audience. Been improving and will keep improving, dunno why there's such a doubt they can become indistinguishable on some level. Lotta research happening on machine learning a lot deeper than image generators
Let’s have a handshake party and wipe the slate clean.
Ya! way to start the morning
I will never use the word AI again.
Totally normal country dood...
It's ridiculous to say that generative AI for images and video hasn't gotten any better in the last year. It is obviously, extremely better at producing coherent and convincing images than at this time last year. That doesn't mean that it's a perfected technology (or that this is indeed useful technology at all), but at least be honest about the pace of development here.
Hi 👋
AI has gotten superficially better in the past year. Chapo is going to feel so silly in another year when the lady who is part bed and part cat looks less like a product of ketamine.
People said that about self driving cars before this. And said that about crypto before that.
You got sucked into the tech hype scam. My condolences
@@Praisethesunson AI and ML predate cryptocurrency. So do transhumanists.
@@Praisethesunson lol read it again. Played yourself.
I would prefer not to
An A.I wrote this comment
An Armenian Islamist.
I think it is more properly called IOI (Inorganic Intelligence). :}
@@inorganicintelligence-IoI IO(/)I is cute bc it spells off/on which is a much more interesting binary than 01
If AI isn’t threatening then why are you so friggin triggered by it lmfao?
yes, indeed. lmfao indeed@@ultravioletiris6241
@@ultravioletiris6241 Capitalism is a threat. Access to speculative capital is why tech companies keep over hyping their trash year over year.
A.I. right now is I'm it's infancy and early childhood. The problem are these psychotic tech bros are treating it wrong like an abusive parent. We all know what happens to children in their teens and early adulthood after living in an abusive household.
Comment For Algorithm
Chapo's takes on tech stuff leave a lot to be desired.
it’s just boring, they’re just saying what everyone has already seen
Oh no it’s a techbro glazer
What do u desire then? Genuinely wanna know
@@Fucyallfr either fresh insight or juicy dirt. not this microwave crap
@@Fucyallfr some critical thinking would be nice. We get Bevis and Butthead style commentary mostly calling Elon Musk names with no thought about society and the possibility of changing it, the rationality of events in the context of capitalism, etc. Also they are constantly saying things that are factually untrue.
This guest is not accurately representing the technical details of these models.
At one point he claims that in generating text or images these modela are just refrencing a huge dataset.
The dataset is not refrenced at runtime. OpenAI could delete all the data they took from the internet to train GPT4 and GPT4 will not be affected even a tiny amount.
You can think of it as a giant stack of numbers that conditionally respond to some context. You pass a bunch of context with some deliverate ommisions past those numbers and check how well the numbers guess what pieces are missing. Then, you continuously nudge all the different numbers in the direction that causes them to make better guesses.
The whole assertion that they cant think or that hallucinations are an integral part of their operation is extremely dubious, especially when spoken with that level of confidence, because we just don't know how the capabilities we see on the surface are actually achieved using that massive stack of numbers that was nudged into performing as well as these models currently do.
To predict perfectly the data they are trained on, they would have to represent the causal processes which originally produced that data. Like, if in the dataset there's a record of a series of scientific demonstrations followed by the observed data, in order for the giant stack of numbers to guess what observations follow from the description of the procedures, it has to reason about the physical processes which caused those numbers to be written down in the "results" section.
More concretely, if you have a description of mixing specific concentrations of acids and bases in a chemistry experiment, followed by a description of the resulting amounts of the produced chemicals, the AI has to be getting better at reasoning about chemistry abstractly in order to better guess those results.
In this interview I see someone who has a take about AI that theyve arrived at because they don't like the attitudes, politics, and general vibes of the people who "cheer" AI on, and then went though a bunch of pop culture sinplifications of what these AI programs do, then took away a misinterpretation of that information which - if it was true - makes the people they dislike look like credulous idiots.
It reminds me of those new age gurus who misunderstood quantum mechanics to mean that consciousness was some sort of immortal property of the universe, and that all minds were entangled together in a great web, and "quantum immortality," "universal consciousness," "minds entangled across all time," etc etc.
They wanted a nice beautiful story about how people are really special and we'll all live forever as cosmic waves or something, and they read a bunch of pop science stories about quantum mechanics and string theory, but didn't bother to actually figure out what the people who wrote those papers meant when they said "the observer collapses the waveform" or "quantum immortality and the many world interpretation."
Similarly, this guy seems to want a story about how tech people are totally wrong about how useful or promising this tech is, and how easily impressed they are by something that is obviously just a flimsy illusion. Then, went and heard some 5 minute UA-cam videos or read a few Time articles, and came away with a misinterpretation of what this technology even does at a basic level.
He even said "average general intelligence" was the thing the field was seeking. Which just isn't what the acronym "AGI" stands for. Thats just not a mistake someone would make if they'd learned enough about this topic that they would be able to make bold categorical statements about how these things work and what they'll fail to achieve by what stage of development.
The damn thing is, I agree that the people in charge of these labs are short sighted idiots who are doing all of this in such a proud, loud, and snobby way because someone called them a naive fool for thinking AI was important or possible back in 2012.
There's a ton here thats completely valid to criticize. I'd point my finger at the fact that these people can't seem to agree that it might be important to be able to agree on where all of this is heading and actually plan in a coordinated and sensible fashion for how to deal with that.
The problem is, not every criticism is going to inspire the exact right policy that we need to take to fix this situation. You can't just say "clearly these people suck in a bunch of ways, why shouldn't I just use whatever argument seems most damaging to them?" The way you pick the take that will actually cause people to think of beneficial improvements to the situation is to try and succeed at making whatever you say actually true.
I know its hard to actually figure out what's true about this technology, because that involves listening to people who are very smug and talk using a bunch of terms that aren't actually common knowledge. But, y'know, kinda notice how a bunch of math weenies are making huge waves in the world, and maybe be a little curious what these weirdos are actually talking about.
I guess this just isn't the crowd that you're supposed to say something like that around? It feels like Matt, Amber, and Virgil are more of the type to go "okay, no, what this person actually means is this, and if you look at it from this angle you can kinda see why that would be true, but they seem to be missing X and Y and Z."
In other words, I predict if I got to whisper in the ear of Chapo Trap House in general and pointed the show at a website like LessWrong or more specifically The Sequences, like for a segement like a reading series, there woukd be a quick scan of some material, then they'd conclude that it's not an easy target to mock nor particularly entertaining to try and comment on. There'd be some sort of reflex that disinclines this group from engaging with material which isn't just so full of nonsense that it's easy to talk disparagingly about it for an hour without running out of things to mock - and not feeling bad about mocking someone for having something like a bad hairstyle, because they're so awful on other topics that it feels like there's a huge budget for spreading that negativity around to other less independently valid areas of criticism.
Am I wrong? Will Chapo turn out to have the correct take here? Does it seem kinda like, on a gut level, that they'll actually miss some trend going on here that'll become important in maybe even less than a year? Do you feel that hint of doubt at the back of your head? Did you sense that quiet note of confusion that sorta flitted around behind your eyes when some of these statements were being made? That a huge fucking alarm bell in my book, it means you can kinda tell this isn't even a totally solid slice of the real story. It's the sensation of seeing a mirage. You want that to be water on the horizon, and you may start frantically lurching towards it, overwhelmed to explosive effort with optimistic certainty that you're interpreting things correctly - but, if you were more aware of what was going on in yohr head when you first widened your eyes at the shimmering sight in the distance, youd have noticed that you didnt even fully believe the story you told youraelf about thay definitely being water. In the back of your head you knew what seeing water felt like, you could tell it wasn't a certainty when you first spotted the shimmer, you may have even had a sinking feeling in your stomach when you started moving towards it.
That's what I think many people will feel when they listen to this. Maybe it's even what Chapo is feeling when they say this stuff. It's a hard sensation to identify and react to, but it's one of those you ignore at your own dire peril.
Loool
Big if true
Spring training starts in two days
Why does it matter what they think? They arent computer scientists or engineers, and as you pointed out - some of the basic misunderstandings from the guest indicate he isn’t very well informed on AI/ML either.
I just don’t think that their observations about the over hype of generative AI will lead to “peril” or however your conclusion ended up
Nah bullshit.
How are they so definitive that this technology isnt going to radically improve in a decade or 2? When I was young we had a lot of these new techs that sucked at launch (motion sensor duck hunt etc.) now theyre so good the biggest selling console was based on motion sensors. VR looked like it sucked for a while but again we can see the Apple headset that by reviews at least suggests theres something here now. I wouldnt remotely be so confident in dismissing tech because were in the duck hunt era.
no one knows what will happen in a decade or two. people are making preposterous claims about what AI will do right now though.
Are we still in the "Duck Hunt era" of the Hyperloop then? Is that just guaranteed to even *exist* in 10 years, let alone outpace existing public transportation methods, simply because that's *your* understanding of how technology develops? Listen to the episode again. They acknowledge the argument that AI is "improving." They are unimpressed. That the Virtual Boy may have been technically superior to Pong does not mean it's an "improvement" over the latter.
@@kingofsting19 No because the hyperloop (elons crap right?) is nonsense, theres nothing innovative there.
They dont remotely acknoledge it beyond a lay up question what do you say to ai improving etc.... which this guy responds with the most insanely confident statements that this stuff is going nowhere.
Its geniunely insane to argue today that you are certain that this technology is always going to be dogshit. Theres an absurd amount of capital flowing into this globally, someone somewhere in over our life is probably going to get this technology to a place that is geniunely impressive to a mass consumer audience.
@@MichaelHachey Yea but ignore those weirdo bluecheck freaks, theyre morons moving off nfts to some new grift, but my take from this guys analysis is basically ai will never replicate something that creators do today in the creative space because of all these flaws. I could have happily argued that in the 90s when I played duck hunt that motion sensors will never get anywhere but they did 20+ years after.
@@Mezzo396 The US military spent $2 billion trying to build the B-2 bomber and it still can't fly in the rain or eject without killing the pilot at least a quarter of the time. It's not "geniuenly" insane to not subscribe to the magical thinking that enough time and money *"must"* correlate to a certain level of technological advancement. Sometimes all you do is push money around and find out why something doesn't work, like the Hyperloop.
Also, who cares what a "mass consumer audience" thinks, bootlick? A mass consumer audience once decided "pet rocks" were a hot item. Come back to me when these AI do something that justifies the amount of money being spent on them, not just be a new thing for the billionaires to speculatively invest in.
This guys sounds like he just got done memorizing "The Power of Positive Thinking." Everything he says is just so uplifting and inspirational! This is what the world needs right now! More positivity! That's the spirit!
As someone who works in AI, the first minutes seemed strange, because one thing AI really brings out for us is that creativity is not a skill. That whatever AI creates has the indelible stains of the mass of culture that produced it. It never seems "fresh" somehow, not without a lot of creative human intervention.
That said, I think everyone is underestimating what could and probably will happen. I'm not a booster of AI. In fact, I hate dealing with it, and I hate computers, too. I'd love nothing more than to see expectations collapse as we hit some fundamental limitation, but they won't. Shit's for real. In a lot of ways, it already really does things the way we do them. There's no fundamental difference between the neural networks in our head, and those in AI software. The only big limitation now is that the AIs have few ways of interacting (and therefore learning about) with the real world, thank god, because robotics has (comparatively) lagged behind and remained expensive. So it only knows the abstract world that's on the internet. It can write a rap about the sequence of operations needed to make a radio circuit with any random collection of components, but it wouldn't know how to even "cut" a wire. But if it was in a robot with good abilities, it could literally figure it out just by looking at videos on the internet.
Yeah man if anyone understands what "creativity" is, it's definitely you, a techie vampire who writes 3 lines of Python a day.
"no fundamental difference between the neural networks in our head and those in AI software" 🤔
I think you wanted a different word than "fundamental"
@@obotopo who hurt you lmfao? Dude doesn’t deserve to be called a vampire just because you’re on a soapbox. Do you really think that anyone with a tech job is incapable of creativity lmfao?
@@ultravioletiris6241Spot the Glenn Dubin internship holder
@@ultravioletiris6241Yes. Lmao. “Tech” and technology are not analogous anymore. Most tech bros hired somebody else to do the creative work that they sell as if they made the advancements themselves. This is why nobody takes the “tech” sector seriously, if you want real creative endeavors in the 21st century, go to local businesses and companies where there are actual stakes or academia.
I REALLY dislike pretty much any form of generative AI. But "it looks bad" is a terrible argument. This tech has obviously made leaps and bounds over the past two years. Frankly, I think you're kidding yourselves about the video generators like SORA. That shit will absolutely fool the average moron on Facebook. I give it a month tops until it's used to generate some kind of horrifying atrocity propaganda.
Nah. It all looks smooth and dead. No soul. You could maybe argue it looks kinda cool aesthetically but it's all just so empty looking.
For sure it will be used for evil though, absolutely.
More right than you know about the last part. Only time will tell about the rest..
yes guys, but look at it from one year ago.
If AI ever does get more advanced and is in android bodies with self awareness I hope there is a civil rights movement for AI autonomy. Yes I read too much scifi, watch Chobits, and play Detroit Become Human.
Thrusting the sentience of a human adult onto something (someone?) seems cruel
Absolutely hilarious that this man talks about how AI makes nothing but regurgitated soulless garbage yet professes to love muse.
so what's your point?
@@drakonian2011 muse are shit.
Classic lud posting. Epic fail.
I can haz cheezburger?? EPIC FAIL! Ok, this wins the Internet ☝️ ROFLCOPTER does the bacon narwhal? this post gave me cancer EPIC WIN
edit: thanks for the gold kind stranger!
@@digitalbear1217 Wow, you really made me think.
@@theenigmadesk Epic fail.
@@theenigmadesk stop when it starts to hurt
Elon bros being stuck in 2011
next time you cover AI, you should probably have a guest that knows what they're talking about.
Found the AI simp
@@Praisethesunson It’s not about being a simp. If it’s important to not fall for the AI hype, it’s equally important not to spread misconceptions the other direction too.
I really love how the "creative class" is shvitzing over this technology. Luddites gonna ludd.
Yeah man not adopting shitty broken technology is just like that time a bunch of workers went into factories and broke the machines. You're so smart.
*Yuh man just have faith it's gonna get so good. Technology gets better so the Juicero is gonna be the juicer of the future maeen.*
I think most people who understand the implications are pretty concerned about this. There's no good that will come from from taking the human element out of creativity.
@@biscuithammer00 Anyone who cries about "taking the hu.an process out of creativity" doesn't understand how AI art or actual creativity works. AI doesn't just make itself, numbnutz. You need a creative human to tell it what to do. AI art cannot exist until I use my human creativity to type a prompt into the text box.
@@abreathingcoffin8089 Yes. Being afraid of the computer because it might take your comissions for drawing furry pr0n is, in fact, exactly like being afraid of cars because they might take your job as a hore drawn carriage driver.
Yea the creative class lost its mind for a long time when modern graphic design (like Adobe products) became widespread.
They're gonna eat their words on this one in a few years. AI is going to become much more capable and prevailing in our lives, probably in a negative way. These guys should really stick to reviewing Kubrick films because they clearly don't know shit about the field of AI.
What are your qualifications for speaking on this
@@abreathingcoffin8089 I'm currently getting a masters degree in electrical engineering and I use machine learning techniques for radio signal analysis. I've also been following the field for years and talk to friends that work directly in AI. That being said, I don't think you need a masters degree to see that the progression from early ML models, GPT-2, GPT-3, chat GPT, and now Sora is continually trending towards something we would consider intelligent, general, and even creative.
☣️Musk-Boy Detected ☣️
@@joshuamarx8209 lol I am absolutely not a musk boy
@@mattcorrigan6639but you sound like you know things your talking about… if anyone here had the knowledge to confirm or understand what you said then we could determine if you actually know what your talking about or just know sone words that fool us…
Israel is closing in on the last Hamas stronghold in Rafah. No word on this yet huh?
How is it the last one? Hamas keeps popping up behind the IDF in "secured" areas
God bless.
What's left to say? We all know how it's going to end.
Oh no!!!!!!! /s
Oh wow, is the country built on forced military service, propped up by the largest weapons-producing country in the world, winning its illegal war against its own minority population, who aren't allowed to leave their fence prison or even fish in their own ocean, let alone have a military?
Epic, bro. You definitely *don't* look like a massive bitch for being excited about that.
Not the best guest
Is this the part where you sit here and deny reality? People can just go watch the videos. You sound like old people. 😂
Did you think that "woman and cat in bed" video looked good? The woman's face is all weird.
Dude commented on a chapo video 20 minutes after it was posted to cry that they didn’t like George Carlin AI hahaha.
They are clearly panicking about computers eating their lunch. It's transparent.
The AI defener has logged on
IMHO people should actually spend some time using the publicly available tools before shitting their pants
You guys sound like luddites trying to sound smart.
what's wrong with being a luddite
They are smart lol
You sound like you didn't actually listen to any of their points, and just want to magically will AI into being good because that's what you'd prefer.
You sound like you've bought stocks in an AI development company
@@NosyFella na, but nice try. Tell me more about things you can't get from a nine word comment.
you guys think humans are magical. that's cute. how quaint.
Your stupid algorithm can't think or be creative, you're just deluding yourself with regurgitations of shit that already existed and was made by people
Hey look, some teenage edgelord who thinks his disdain for his fellow man makes him a freethinking genius and not a boring cliche just looking for an excuse to not contribute to the world.
another thread where everyone knows each other. Can you guys take this into the kitchen please?
-Posted by Cypher from the Matrix
You think ai will become God so uh not really an own
god, felix is so unfunny
Why do you think God cares about who makes you laugh? Stop talking to him, he’s never going to answer.
“Joining Felix and I today…” Sorry, you really need to learn English grammar. I’m not even a native speaker, and I still know it’s wrong.
You have a real late 1930s German approach to grammar
English grammar is what we Americans say it be.
It is correct. There are many different ways to say that expression, and this one is fine. Besides, native speakers know the difference between spoken word and written word, and since English is notoriously complicated, much leeway is given to native speakers.
Literally proper grammar. Will speaks with a half NY/half mid-Atlantic accent and you don’t think that half his time in primary school was just getting wrapped on the knuckles for improper grammar? Come one jack
literally who cares
THATS MY MUTUAL IN THE THUMBNAIL @L404
what frustrates me is these AI video tools obviously cannot do any of the pointless hollywood fanservice stuff these people say they will, but they could absolutely be used to make some incredible experimental stuff, we're just not seeing a whole lot of that yet. one incredible example of the potential is this michael jackson film, beautiful haunting work of art that uses AI to great effect, exactly the kind of thing we should be seeing more of: ua-cam.com/video/P2zPO-MFBkQ/v-deo.htmlsi=PlPNPfEty7f29eQ1
Yup the marketing and hype is definitely over the top. Tech does this a lot for various reasons such as to seek investments or customers, and now you get to throw in even more transhumanist mysticism than usual.