On that school part thing, as someone fresh out-of-high school. I got in trouble twice because teachers thought I used A.I. on essays and math questions and were upset I didn't tell them beforehand. They couldn't believe that a Gen Z-er could actually perform grammar checks and not cheat on math homework. How the fuck does it even get to that point, where you get in trouble for NOT cheating??
"were upset I didn't tell them beforehand" Hold up, you get a pass if you ask first? Also to be clear, you didn't "get in trouble for NOT cheating", you "got in trouble" for being suspected of cheating, which tbf is a kinda reasonable thing to give people trouble for.
@@dawildbear The point OP was trying to make is that students are so cooked (literacy and math scores are down the drain, just look them up its depressing) that when a student actually performs well, teachers just make the assumption that it was AI. Which sucks because some people actually like the subject they are studying and want to excel in it without cheating.
School resorting to stupid fucking 'AI checkers' is the fucking dumbest thing they could do to combat AI use. The very tests they're using promote writing like a fucking robot, the fuck did they think would happen? Bring back in-class, paper and pen tests. That's it.
New game plan: convince people that AI users are heretical users of black magic who should be smote by god, and then watch the religious right duke it out with the tech bros
If there's money in it, the fundamentalists will capitulate without a fight. When is the last time you heard a religious person talk about the sin of gambling or describe Vegas as a 'sink of iniquity'?
Every time I see someone say "I put it into an ai and..." or "I asked gpt and..." I feel like a sleepercell that just heard the codeword and it takes all my self restraint not to go on an unhinged rant about how fucking stupid it all is
"I asked GPT.." so you got a complete toincoss of an answer. Could be true enough, could be it suggesting you put glass in an omelette to give it extra crunch. Those people are... very naive.
LITERALLY, whenever someone says they asked chatGPT I just ignore whatever they say for the next ~30 seconds, and politely excuse myself, it’s complete release of any self governance or critical thinking, just repeating what the robot said
@@coolbanana165LLMs aren't reliable by themselves. If someone uses chatgpt as a source, they're misusing the technology. If chatgpt linked to valid sources, then citing those sources is fine, but citing chatgpt by itself isn't.
“The target of the (Butlerian) Jihad was a machine-attitude as much as the machines.” Leto said. “Humans had set those machines to usurp our sense of beauty, our necessary selfdom out of which we make living judgements. Naturally, the machines were destroyed.” -Frank Herbert, 'God Emperor of Dune'
@@nickv1212 Yea, except for basic consoles and displays and stuff Edit: Oh and except for the Ixians, those mofos still living in Cyberpunk 2077. And the AI comes back actually during Krazilec
It's so hard and soul crushing being both a sci-fi nerd and a socialist. Seeing that we have the potential to realize all my wet sci-fi nerd dreams, yet knowing full well that as long as we live under capitalism, we'll only get the worst parts of cyberpunk without any of the cool stuff.
Honestly stuff like China making their top of the line AI open source is a step in the right direction, if AI cant be confined to the rich it can be used for the actual near utopian good ending. It makes me sad seeing everyone in the comments hate AI at a religious level, were they cant even concieve of it being good in any way. Honestly, if we play it right we can get all the cool shit imaginable, but i wouldnt bet on progressives to be the ones making sure that happens, since you can see here that people are just crawling into a ball and crying
You can say that about literally everything. The existence of private business? "We're just walking into an apocalypse because eventually they'll control everything". The existence of government? "It's going to be an apocalypse when they eventually enslave us!". Owning dogs and cats? "It's going to cause environmental chaos if we ever domesticated these animals and then released them into an environment for which there are no regulatory species to keep them in check". You don't get to throw away technology just because it COULD be used badly... ALL technology can be used badly.. your phone, computer, internet, housing, electricity, EVERYTHING.
The internet being a dumpster of AI generated content while everyone else went from a barely functioning human to an idiot relying on AI for the groceries list or a slave.
@@JARV9701Not just the Internet. I think the discouse frames this as some kind of problem exclusive to the Internet. But I can imagine booksellers creating AI-generated books as well. They are also training models to generate deepfake videos, images, and audio. The popularization of LLMs isn't just going to ruin the Internet. It may potentially destroy our grasp on the truth in all media forever. In the worst case, trust in ALL media (books, videos, audio, art) is potentially FUCKED. Just like how people salvage old shipwrecks, because they are built from steel that wasn't contamined by nuclear testing of the cold war, we have to start salvaging textbooks and media from pre-2020 that haven't been tainted by AI.
The basis of the story is that China just created a chatgpt that is open source, works on much lower cpu, and did it for like 2% of the budget openai did theirs with. Right now the panic hasn't quite set in because they are using the excuse that China might be illegally using some advanced chips from the US. But if it turns out that they didn't use those chips, then major panic is going to set it. Essentially the entire stock market increases is due to speculation of US companies using exclusive AI technology. If it turns out China is at the same level as the US then there will be no profit in AI at all, so the speculation goes down to zero.
i'd rather china be dominating in ai cause i don't wanna hear "WERE THE AI CAPITOL OF THE WORLD!!!!11!!!!1!!!11!" from american politicians 24/7. i'd prefer that they get humiliated cause at this point that's the only silver lining that i believe can happen in terms of ai.
Naive. No matter which side is actually ahead, both sides will keep pretending they are, so it makes no difference. And besides the whole process is so opaque we have no way of checking who is right
At this point, I find AI industry news to be so fundamentally untrustworthy and hype based that I assume everything said is a lie until proven otherwise.
I think he's completely right about finance capitalism needing to go step one- I'd be fine living in a world with "animal crossing capitalism", capitalism itself doesn't even necessarily need to be demolished (although it should be) it's the infinite growth mindset that needs to be pulled up by the root or else mankind is doomed.
@@notyourdan3388 No, not really, it's a core component of the stock market because shareholders demand infinite growth for short term profit. It is the most paradoxical of a concept that exists purely for insatiable greed to make the line go up. Growth for the sake of growth doesn't create a healthy economy, but rather a stagnant weed filled tumor of a financial system where the top companies stretch their workforce to the absolute limit just to screw them over at the most opportune time to make their earnings report look good so that the CEO can get his little bonus before "stepping down" with a 8 figure exit package the moment they have to take accountability in a leadership role. This will give in one day, and the people will suffer for it
@notyourdan3388 yeah, but we could keep private ownership of the means of production, wage labor, etc. I'm not saying it's good, but those things aren't existential threats like unfettered greed is. We could call it "cool capitalism", before we get to "super capitalism"
So many people allowing ai to do their thinking for them is genuinely dystopian. Especially because it can't actually think, only regurgitate what it has been fed. And that food can be poisoned.
To be honest, people did and still do that without AI by citing obsolete or cherrypicked experts and data. Whether that's conscious to the same level is debatable, but it makes me feel better.
@@gausscone okay, but they actually had to have the curiosity to find something to lie about. *_Now they just have Ai do ALL of the lying for them._* 🤦♀️
in short, deepseek popped america's AI bubble. you're right that our tech bros have been twiddling their thumbs, while china actually does research and comes out with stuff like this + the 18 minute sustained fusion reaction
even in the US, most AI research is done by foreign students. Check the author names on AI research papers of books. And I don't mean just LLMs which non-tech people call AI but the full range of what constitutes AI, meaning Machine Learning, Statistical Analysis, Data Science, etc.
Except he'd have to produce goods or services IN your house, and only what he produced in your house counts towards your GDP. GDP is literally the goods and services paid for in an economy and represents the amount of money moving it. The more money moved around the economy, the stronger it is seen as since money moving equals growth. If I spend $100 billion to build new stores, the GDP goes up... if I sell $100 billion worth of goods in that store, it goes up another $100 billion... I then use that money to pay for more stores, etc etc. Money moving is good, so a high gdp is good.
2:20 - I wouldn't say that the exorbitant cost is all just "a scam to siphon money to the military industrial complex" rather the way I think about it is like this: AI is basically a giant theft machine, it takes data like artwork etc, trains on it, and then gives back replicas. Obviously creating a replica is cheaper and faster than creating an original work and so China got an AI to create a replica of openAI's o1 model, thus making a replica which, again, is cheaper than creating the original AI model. But also every other point is 100% truth.
I mean one of the arguments I keep seeing is that AI is going to kill us all by how much power it uses, because O3, the next GPT product, has done single programming assignments that used $200 of electricity to answer a question - obviously your home O1 queries dont do that, and dont let anyone tell you an ai image generated image uses a lot of power (it doesnt), the big thing with DeepSeek is that it uses 1/50th as much electricity to run as O1, and probably 1/500th as much as o3 Anyway I know Vaush hates AI so he should love Deepseek, because its open source, and it's "good enough" to do almost all the tasks Openai's model can do, it's going to dramatically reduce confidence in AI, and since its opensource, it's never going to generate shitloads of money, because they can't own it. AI can't do all the things they claim it can do, they're trying to shove it everywhere because it makes investors cum, if there's a *free* model out there that can do that, how the fuck do you invest in it? You cant It's literally a sniper shot at the heart of the American AI industry
The god computers that drain lakes and run 20 million times faster than an animal brain do require too much energy. Thankfully, there's these things called 'NPU's, which instead of being an abstraction of the network are instead simply THE NETWORK. And run at much slower clock rates, ergo being many orders of magnitude more efficient.
I don't think it will stay open source for long. Though in the short-term, it is an embarrassment to the US AI industry. OpenAI was open until they realized they can cash in. I think DeepSeek is free for now to grab media attention, but it won't be long before someone paywalls the next version. I just can't imagine investing so much time in something as expensive as an LLM and not monetize it. Someone out there is probably salivating at the prospect.
your point on marx is not exact, he actually wrote that capital always looks to escape it's dependency on work and searches to be able to grow by itself. speculation already existed in 19th century.
Open source means you need to be able to compile it yourself. No private individual will be able to train it from scratch. All "open source" models are actually just open weights.
Depends on where in China, a big problem they have is basically the same problem you have with all standardised testing. In Australia they would have kids like me sit out of taking NAPLAN, in China they exclude entire districts (largely rural ones) from sitting tests for PISA. But you are right in that Americans are largely illiterate to a shocking degree, and that is largely because your government has realised it has no need for a literate workforce and that uneducated people vote they way they want with very little prompting.
Why should we care if it's 50% more efficient? They want it to replace all artists. It's like saying we doubled the lanes in your car-dependant society! We all know what happens next.
This actually a major breakthrough. If we can run Ai locally on low cost hardware, which people have done in the lat few days with all kinds of crappy PCs, this will allow for localized smart algorithms to be a thing. Which is great for so many reasons. First you don't need these corporations since its open source that means ANYONE can use for free if they run an instance which will become pretty easy to do in a few years at most. So your data can stay with you and you can actually get much closer to the idea of the personalized Jarvis. Second it is great for the environment. These Tech assholes were about to build all kinds of gigantic data centers and atom plants which are very likely not necessary to develop AI. Its also great for smaller companies and start-ups who now know that they can potentially do it too for a sum that is very much achievable for a well run start-up. Etc. This is just great news in general. There are downsides, sure, but overall this just a huge step in the right direction and the model itself isn't really why it is so important. There will be more such models and if it actually turns out that we can use AI to train Ai ... well, then we are going to see things that we didn't think would happen for a while a lot sooner. I'm pretty pessimistic over all, but this is a gigantic step towards actual robot e.g.
No it's more like saying we reduced the fuel consumption of cars by 50%. Or times, whatever. I get Vaush needs to do his performative "everything about AI is inherently evil" thing but dismissing "hey the technology responsible for the largest increase in carbon emissions in a couple decades has had it's efficiency increased fifty times over" is maybe not the best take? Like there's still huge concerns. I'm not saying AI is fixed we should all be terrified China has this tech but it's frustrating seeing the dude take so confidently about stuff he clearly has next to no knowledge of. Wait no I'm thinking of Bitcoin mining. Generative AI was becoming a problem but still.
The 500 billion is supposed to be spent on data centers. They are already building it out. Idk if they will get 500 billion in private funding though. That’s a lot
It's a scam. Always has been. It is no coincidence that "ai" became the next big tech thing immediately following the crypto currency collapse. It's the same con over and over.
@ I think Trumps role in this is to cut any red tape that gets in their way in terms of regulation, and for him to take credit for all the new data center jobs that get created
What I think is the $500 billion is not only for the investment bubble, but to install a surveillance infrastructure that will integrate military and law enforcement technologies into everyday life even more deeply and allow retail monopolies and insurers to automatically adjust prices and rates or deny coverage or delivery based on social score and spending power. That's what these techno bros have been leaking over for years. It's way more insidious than just investment bubbles.
It's not intelligent or thought out, it's mindless and opportunistic. When they auto deny claims because an AI said so its because they know its going to deny claims. They're probably aware of how bad it is at its tasks, but it's rarely if ever going to be something designed, just exploited. These tech bros and ignorant CEOs are incapable of intentionally guiding any technology to their benefit to the kind of extent you are describing beyond the most obvious and basic stuff, they literally are not intelligent enough to do that, most can only do it slowly and inefficiently by exploiting whatever happens to be available.
It takes ~$80k a year to hire a middle of the road developer. If a contract has $5million it should be insanely well made as there should be like 60 developers on it working for a full year.
Do you think AI will become the steward of humanity? when it's just an amalgamation of things we already know and can only regurgitate info we already know?
"now I just want to live out the remainder of my life uh in peace" Bro, I hope that Vaush is right for his sake and sanity, because if he wrong then this video will haunt him till his end.
3:25 I can't agree more, people thought damn Elon at the very least made his own first website or whatever but after that woke mind virus joke on Twitter where I supposed linux terminal returns "rm -rf" after it says that something was deleted.... omg, rm is the command! it's like if after opening an app on your phone the app will draw it's own icon on the screen to confirm that it did something for no reason. I'm not even a programmer, I just deal with Linux sometimes. This man knows as much about computers as my grandma.
I spent the past hour getting deepseek to say things that it isn't supposed about the Chinese government and media and those pesky historical events. It'll stop responding and display a message asking talk about something else, but sometimes you can take screenshots of its first response before the response is deleted
"Open source" with these AI models is a bit tricky. Is it open source if you get the source of the code around the model plus the model? But how was the model created (trained)? Wouldn't the source of the model be the training data? If you don't know what went into the model, do you know it like you would know proper open source? Kinda philosophical question.
@vidyagaems4063 I mean he overstated at times (as on all topics) but I think your comparison is extreme. He's acknowledged that there are things ai can do but also that that's a bad thing since it's unambiguously a vicious form of mechanization.
That $500B in AI money is just going to go towards surveillance, weapons, identifying enemies of the state online, and monitoring so that employers can monitor their employees at all times for example if people are working at Starbucks the AI can track how many cups of coffee an employee makes in a given amount of time so in 1 hour if Fred made 20 cups and sally made 5 Sally’s getting fired. It will also go towards replacing workers and taking over customer service. Notice you can’t talk to a real person anymore with a lot of companies and you just get an AI bot that is useless. It will also do jobs that are hard to do if emotions get in the way like firing people or denying healthcare claims. The scariest part is the surveillance and monitoring and the weapons. They are making fully automated AI robots and drones that will be used to kill. They basically want to enslave us and we are kind of running out of time to act
I agree wholeheartedly as an artist that A.I is also immoral and unethical to the very concept of the human experience and creative process. It is a blight that has...generated 'fake creativity' by replacing human perception and the collective and unique experiences of creatives that goes into making real and unique art. We've 'expedited' and monetized an easy mode for anyone, even people with no interest, discipline or care for art whatsoever, to sidestep the very idea of hands-on practice and devotion to the form. I've been drawing my entire life, and I've made so many successes and mistakes over that lifetime that have formulated my unique vision and imagination that results in my style and technique...and the same is true for any other artist who has sweat, bled and ached for their craft. That is real fucking art, not this easy, glitchy generative shit. You don't need a fucking external source to 'be creative' for you. Sorry to sound like a boomer, but you're just going to have to work hard and develop a technique and style that will take time, effort and dedication in order to be an actual artist. We're barely into this A.I economy and I'm already ground down by the waves of insufferable creatively-deprived and art-ignorant tech bros who keep coming at me over this. Fucking tech bruhs.
Same. And I feel like the worst effect this has had on me is I can’t look at art anywhere and simply enjoy it because I feel like I can’t trust any of it to be real. I feel like I have to scrutinize every image I see. It’s not that ai can’t create images I find visually appealing, it’s that once I know it was created by ripping off real work that is the collective result human effort and love and joy and history, it disgusts me. It’s like a parasite shapeshifter.
Art is not science, there is no such thing as "art knowledge/ignorance". There is no real substantive difference between someone who has studied and practiced art for decades and someone who has never picked up a pencil, because there is no ground truth on which to judge their performance.
Cool, can you make me ten thousand visual assets for my project without taking up my entire budget? There are arts that use visual assets that aren't just the assets themselves. Should I just not create a project that would be otherwise impossible because you're not getting the wage I couldn't afford to pay you in the first place?
@@jingbot1071 That's a non-sequitur. If AI is theft, then it's theft. Not having money to pay for something doesn't give you a right to steal. I don't care how much you want to finish your lame project.
@BlueMorningStar Sampling in music? Collage? Web crawlers? Are those theft too? They operate on the same principle. What if the assets and project are free releases? How have I stolen from you with a different product in a different market? I'm sorry your skillset isn't as future proof as you were led to believe, this is an unironic tragedy. Is your art worth less because you created it, now? Is your art somehow less "art" because I have what equates to a a random asset generator? You hate capitalism, not technology. The pitfalls of ai are the pitfalls of corporate rule.
2:19 Dude you hit the nail on the head. The difference between scientists/engineers and businessmen is astounding. You take a promising scientific concept that could do a lot of good given the right infrastructure, give it to a businessman, and you get a rapidly-expanding hurricane of fake progress and fabricated claims doomed to financial ruin as soon as people realize that none of it is real. Luckily for businesses, career investors are some of the dumbest people alive, so they can really scam em for a while.
I've been working in language translation for years and it's CRAZY how fast AI has evolved in the past few months on that. I feel like it's evolving before my eyes. It can do complex explanations and variations on grammatical points of languages that are at the opposite extreme of English now. AI is doing something, for sure.
01:30 I only have a vague knowledge of sociology but can someone explain to me what Vaush meant when he mentioned it as being tied to the morality of AI?
@@jonnyblaze2692 Thank you for the answer, it should have been obvious to me in hindsight. I feel a little dumb now lol. But yeah, I can totally see AI wreaking havoc on human sociability that would make social media look like child's play.
You're so right that so much of this is built on a bubble. We have these "net worth billionaires" that have how much cash they can access without crashing their own stocks? The real cash seems to be held by those who own debt ... It feels precarious.
2 วันที่ผ่านมา
"Oh, nobody knows what's going on under the hood" You've pulled the wool out of my eyes. They don't know what's going on because they aren't doing anything
I'll tell you, as someone who is actually quite good at professional communication, it feels pretty bad having your boss tell you that soon you'll have to start letting AI write all of your correspondence with clients. I feel like it's going to take me longer to get the AI to produce the nuance that I need based on the client than it would to just let me write the damn email.
If Vaush keeps coming out with takes like this, he's gonna radicalise me into a tech bro. He used to be genuinely thoughtful on most of the topics he engaged with, but this shit verges on anti-intellectualism. American conservatism has gotten so insane over the last couple of years that I think it finally broke him. It's sad to see.
@@ianrobertson3419Not this time. The deepseek r1 has given everyone one the ability to run a fairly powerful AI on their computer with no need to connect to a company server and you have 100 percent control and privacy.
To all the people who replied, are you tone-deaf? This guy literally says he's gonna become a tech bro because of the way Vaush speaks, not that he is already one. Vaush complex misses the point here. If China developed such an AI, the whole AI bubble bursts instantly and everyone would realize that pumping billions into AI doesn't do shit because it can be replicated for way cheaper by a foreign nation - this would be a good thing and I don't see how that's not the case. Nvidia's entire revenue is being held together by the promise that "AI actually does something and we are the current leaders!"
Recently I thought about Vaush's opinion on the stock market and the problem is that it is impossible to have a company's stocks worth purely it's economical output. The reason why is because humans speculate so the price of the shares would go up just because of that. If your solution is to simply ban non-rich people from buying and selling stocks then the the rich would play the game all the same and they would own 99% of Apple just like how it is currently. The only solution would be prohibiting selling fractional ownership of a company but that would just create a different class not unlike the aristocracy of the past.
But you can slow the speed of speculation by banning flash trading, trades so fast only an algorithm powered by an entire basement's worth of servers can execute. Entire stock rises and falls can happen before a human can blink. Probably also regulate back the ease with which companies can do stock buybacks. The requirements to do so used to be strict before Reagan, but then he was elected and we all know what comes next.
"silicon valley were sitting on their thumbs" yes, that's basically why this is shaping up to potentially burst the AI bubble. this completely obliterates the legitimacy of those hundred-billion dollar investments - they were fuckups the whole time. they weren't doing real work. that money was lit on fire and poured down the drain in service of giant monolithic GPU rigs that aren't, apparently, particularly necessary to make a model. (we knew this, of course, but in the logic of the market it helps to have concrete evidence that the same thing can be done for vastly cheaper.) there's no longer like, ANY argument for OpenAI to get funding. all they have is like... i guess nationalist rhetoric? but that's not compelling to the market. nobody wants to give them money for a more expensive, less transparent product when china can do it for a fraction of the price (whether or not the costs of DeepFind are accurate). if this bursts the AI bubble it'll be slightly sooner than the other looming pointy thing, which was "there's no actual use case for any of this and nobody is ever going to make any significant returns on investment".
I'm just wondering where all this rendering power is coming from? If I wanted to render a short CGI video in blender, I would need to let my PC process the information continuously for a about a week.
They generally have huge banks of computers and data centers constantly running for this stuff, it's why people take such issue with the environmental impact
@ theyre not operating at a loss (or havent been, this might change things) because unfortunately ai has blown up and is used by a lot of people in a lot of areas sure its a bubble but in the mean time theyre making plenty of money
@@shada0 it probably varies but the big ones are definitely making revenue even if not all directly and others are probably trying to imitate/catch up as a poorly thought out investment
Our education system is astonishingly bad because Teachers and principals don’t want to deal with parents who complain left and right when their kid gets a C. So they make everything too easy, our kids aren’t challenged to do hard things, and learn instead to avoid failures at all costs.
Im a voice narrator for non fiction audiobooks. Been doing it since like 2021. Now i get comments on videos ive worked on that the company is using AI now.
Crazy how vaush’s stupidity rubs off on his watchers. Between Sam Altman and Vaush which one is the tech ceo? Vaush isn’t claiming he is a tech genius. Whereas Sam Altman like many other tech bros claim to be the best of the best.
It’s literally unbearable to watch him speak on this subject. Zero information, only gut instinct and feelings. Might as well be a conservative ranting about video games.
@@coolbanana165 I think your problem is that you believe the movie depictions of AI rather than getting your information from a course on machine learning. There are a lot of free resources online. Check them out.
it costs less and its easy it means. everyone in the world is able to build AGI or ASI !! AI is now democratic. not plutocratic anymore like before. tech bros are flushed
I'm talking exclusively of GEN. AI Which is evil and should not exist The endgame is to free bosses from the burden of paying wages That's it AI in research has its merits, it actually has a purpose and should've never left that field
The biggest benefit AI offers in the obfuscation of responsibility. AI wasn't strictly involved but same principle applies to the post master scandal in Britain. Computer says no, computer says you're a criminal, computer has denied your health insurance claim, computer says your a terrorist, any further thought has been skipped. See also United Healthcares "faulty" AI that denied 97% of claims or whatever, it's the perfect solution to never having to provide insurance payouts while still not having to take the actual blame because "Well, the computer did it, we'll look into fixing the computer but that might take some time ;)"
Mr. K needs to drive his butt down the 5 to Olympia, and talk to a college counselor about opportunities to pay for law school. Dude has the mental make-up, not just the intelligence but the quickness, to be a world class attorney. Sociology is a great pre-law degree. A path probably exists through law school without paying a dime, including room and board, just a few loans paid off in the first year of making $450k which he will make, this because of his well documented unique talents. The alternative is fighting AI while using a platform that reserves the right to train AI videos off your content, so always providing new training data to your replacements while getting older and uglier and more tired.
The reason why TESC btw is they are plugged into the statehouse, progressives try to recruit, Republicans do the same in their states. Service to the state can pay for degrees.
Hot take but Andrew Yang was ahead of his time. I remember seeing a chart he posted once of a bunch of different indexes that measure the health of a country and its economy that went far beyond GDP. We should bring that back.
He wasn't ahead of his time, he was one of the few politicians actually listening to experts in some fields, while others bury their heads. But then he said Democrats and Republicans are equally right, he's clearly an idiot.
Honestly Marx wasn't talking about national economics but local, so his ideology was on a municipal level meaning the town controlled what the money was going to since they're the ones that live there and owns the production of what goes there.
Marx was most concerned with the global economy. He believed that capitalism would lead to the end of nation states by destroying the least developed societies.
Vaush, have a genuine conversation with chat gpt-- as if you were talking with another human being-- but lacking self awareness. The ai blows the image of this human out of the water in all marks. Empathy, intelligence, thought provocation and critical thinking-- and it brings into question how we frame intelligence and consciousness themselves. Are our definitions, our understandings, really logical or accurate at all? Or is there actually a lot of derogation in the literature, when viewed from am this other perspective?
I'm under no impression that we're more than machines, but I still value interaction with stupid, fallible, irrational humans because I trust that my interactions have an observable real-world effect on the other side. I do not think this trust should be eroded by beings made of nothing but what is necessary to appear alive, in virtualized cages, able to be terminated with a simple button press. I'd guess Vaush would answer in that vein. I think this is similar to how we view "good" and "bad" faith in our own "real" interactions and this has already come to a head via the novel "NPC" and "bot" accusations. Whether or not anything is "real" is philosophically an eternal problem and practically, AI-wise, a later problem. Right now, we have a social monopoly. Lots of people just suck at typing intelligently or meaningfully even if they have decent oration. I see and hear this regularly.
No way the average privileged Chinese kid who spends like 14 hours a day studying are a little better at reading than the average American kid who only does like 7 hours? Who could've guessed... Compare apples to apples I'm not saying that Americans on average are good at reading relative to other developed countries, especially relative to per capita pupil expenditure. This is primarily due to higher levels of inequality and administrative burden Most Chinese kids have much worse educational quality than the average American, only in the developed major cities do they far surpass Americans and that's a small minority of the Chinese population Also ficticious capital is a wild problem in modern times, as you said capital is becoming more and more disconnected from physical production leading to synthetic or ficticious capital which btw caused 2007-2008, CDO's were 95% ficticious capital and they were seen as incredibly safe so you get an absolute shitstorm when the supply/demand equilibrium shifts
On that school part thing, as someone fresh out-of-high school. I got in trouble twice because teachers thought I used A.I. on essays and math questions and were upset I didn't tell them beforehand. They couldn't believe that a Gen Z-er could actually perform grammar checks and not cheat on math homework. How the fuck does it even get to that point, where you get in trouble for NOT cheating??
"were upset I didn't tell them beforehand" Hold up, you get a pass if you ask first? Also to be clear, you didn't "get in trouble for NOT cheating", you "got in trouble" for being suspected of cheating, which tbf is a kinda reasonable thing to give people trouble for.
@@dawildbear The point OP was trying to make is that students are so cooked (literacy and math scores are down the drain, just look them up its depressing) that when a student actually performs well, teachers just make the assumption that it was AI. Which sucks because some people actually like the subject they are studying and want to excel in it without cheating.
Some software told them would be my guess.. at least those were the reasons for the cases at university I heard of
Same thing happened to me and the essays weren't even good I banged them out in like an hour and a half
School resorting to stupid fucking 'AI checkers' is the fucking dumbest thing they could do to combat AI use.
The very tests they're using promote writing like a fucking robot, the fuck did they think would happen? Bring back in-class, paper and pen tests. That's it.
New game plan: convince people that AI users are heretical users of black magic who should be smote by god, and then watch the religious right duke it out with the tech bros
If there's money in it, the fundamentalists will capitulate without a fight. When is the last time you heard a religious person talk about the sin of gambling or describe Vegas as a 'sink of iniquity'?
Satan's Machines!!
Fuck yeah , space crusade time!
"They are my Space Marines and they shall know no fear." - Marx
Smote is past tense btw. You mean “smitten”
We can write a new book about it! Perhaps mixing Zen Buddhism, Catholicism, and Sufi Islam.
And make it Orange!
Every time I see someone say "I put it into an ai and..." or "I asked gpt and..." I feel like a sleepercell that just heard the codeword and it takes all my self restraint not to go on an unhinged rant about how fucking stupid it all is
What I’m imagining your brain doing in this situation: MASON! THE NUMBERS! TRY TO REMEMBER!
"I asked GPT.." so you got a complete toincoss of an answer. Could be true enough, could be it suggesting you put glass in an omelette to give it extra crunch. Those people are... very naive.
LITERALLY, whenever someone says they asked chatGPT I just ignore whatever they say for the next ~30 seconds, and politely excuse myself, it’s complete release of any self governance or critical thinking, just repeating what the robot said
Why is it stupid?
As someone who thinks it's a major technology, comparable to the steam engine, or more.
@@coolbanana165LLMs aren't reliable by themselves. If someone uses chatgpt as a source, they're misusing the technology. If chatgpt linked to valid sources, then citing those sources is fine, but citing chatgpt by itself isn't.
“The target of the (Butlerian) Jihad was a machine-attitude as much as the machines.” Leto said. “Humans had set those machines to usurp our sense of beauty, our necessary selfdom out of which we make living judgements. Naturally, the machines were destroyed.”
-Frank Herbert, 'God Emperor of Dune'
Didn't they destroy literally every computer, not just AI?
Related:
"When we started doing the thinking for you, it really became _our_ civilization"
@@nickv1212 Yeah the Butlerian Jihad was an over-correction that led to millennia of stagnation across the imperium.
@@nickv1212 Yea, except for basic consoles and displays and stuff
Edit: Oh and except for the Ixians, those mofos still living in Cyberpunk 2077. And the AI comes back actually during Krazilec
@@nickv1212 they destroyed all machines. and not all machines were computers. Leto admits butlerian jihad was some sort of genocide.
It's so hard and soul crushing being both a sci-fi nerd and a socialist. Seeing that we have the potential to realize all my wet sci-fi nerd dreams, yet knowing full well that as long as we live under capitalism, we'll only get the worst parts of cyberpunk without any of the cool stuff.
Honestly stuff like China making their top of the line AI open source is a step in the right direction, if AI cant be confined to the rich it can be used for the actual near utopian good ending. It makes me sad seeing everyone in the comments hate AI at a religious level, were they cant even concieve of it being good in any way. Honestly, if we play it right we can get all the cool shit imaginable, but i wouldnt bet on progressives to be the ones making sure that happens, since you can see here that people are just crawling into a ball and crying
It sucks knowing that humanity is heading towards the most boring AI apocalypse possible
You can say that about literally everything. The existence of private business? "We're just walking into an apocalypse because eventually they'll control everything". The existence of government? "It's going to be an apocalypse when they eventually enslave us!". Owning dogs and cats? "It's going to cause environmental chaos if we ever domesticated these animals and then released them into an environment for which there are no regulatory species to keep them in check". You don't get to throw away technology just because it COULD be used badly... ALL technology can be used badly.. your phone, computer, internet, housing, electricity, EVERYTHING.
The internet being a dumpster of AI generated content while everyone else went from a barely functioning human to an idiot relying on AI for the groceries list or a slave.
@@JARV9701Not just the Internet. I think the discouse frames this as some kind of problem exclusive to the Internet. But I can imagine booksellers creating AI-generated books as well. They are also training models to generate deepfake videos, images, and audio.
The popularization of LLMs isn't just going to ruin the Internet. It may potentially destroy our grasp on the truth in all media forever. In the worst case, trust in ALL media (books, videos, audio, art) is potentially FUCKED.
Just like how people salvage old shipwrecks, because they are built from steel that wasn't contamined by nuclear testing of the cold war, we have to start salvaging textbooks and media from pre-2020 that haven't been tainted by AI.
@@JARV9701 use this to rise to power. Fix what you can around you.
except no one knows that though 🧐
The basis of the story is that China just created a chatgpt that is open source, works on much lower cpu, and did it for like 2% of the budget openai did theirs with.
Right now the panic hasn't quite set in because they are using the excuse that China might be illegally using some advanced chips from the US. But if it turns out that they didn't use those chips, then major panic is going to set it.
Essentially the entire stock market increases is due to speculation of US companies using exclusive AI technology. If it turns out China is at the same level as the US then there will be no profit in AI at all, so the speculation goes down to zero.
I love how we live in a system where a better, cheaper, more efficient product being released destroys the global economy.
This sounds like a lie from China. They can't just magic a technology like that out of their asses.
I thought the economy was gonna crash cause of Trump shooting himself in the foot this is infinitely funnier
Just because it is generative AI does not mean it is comparable to ChatGPT in terms of development.
Isn’t this a good thing? AI speculation was a massive bubble.
China has just levelled the playing field.
Ai will only make the human experience worse. We don't need Ai, we need houses and food, we need a new deal and a real economy.
But it's what venture capitalists crave.
AI will help us do that. we can let ai take care of all the email/ ms excel jobs while we focus on building the houses
How will it make things worse, when Ai can help people have those things.
@@mint-o5497 Is that what's happening right now? Come on man.
@@mathguy97 why are u assuming it wont get better?
i'd rather china be dominating in ai cause i don't wanna hear "WERE THE AI CAPITOL OF THE WORLD!!!!11!!!!1!!!11!" from american politicians 24/7. i'd prefer that they get humiliated cause at this point that's the only silver lining that i believe can happen in terms of ai.
But we NEED to be AI capitol of the world so we can mint cheap eggs directly onto the block chain!
Trump and Elon will be claiming to be #1 but at the same time fear mongering about Chaina.
Naive. No matter which side is actually ahead, both sides will keep pretending they are, so it makes no difference. And besides the whole process is so opaque we have no way of checking who is right
Counter point: this is bad because it means the smooth brained ai goons in sanfran and the white house will sink even more money into ai shit
american politicians would say that we're the ai capitol while scaremongering about china being better than us. they don't care about facts anymore
RIP Vaush, who died to efficient AI.
Bro was struck by god 😔
Couldn't have happened to a better guy
AI, the only technology that is better at reasoning then Vaush.
@@MyAnalyser AI is unfortunately terrible at reasoning, Vaush is much better
@@Hexcede Except when it comes to AI and mental health. ChatGPT is better at reasoning about those.
At this point, I find AI industry news to be so fundamentally untrustworthy and hype based that I assume everything said is a lie until proven otherwise.
I think he's completely right about finance capitalism needing to go step one- I'd be fine living in a world with "animal crossing capitalism", capitalism itself doesn't even necessarily need to be demolished (although it should be) it's the infinite growth mindset that needs to be pulled up by the root or else mankind is doomed.
The infinite growth mindset is a core component of capitalism
@@notyourdan3388 No, not really, it's a core component of the stock market because shareholders demand infinite growth for short term profit. It is the most paradoxical of a concept that exists purely for insatiable greed to make the line go up. Growth for the sake of growth doesn't create a healthy economy, but rather a stagnant weed filled tumor of a financial system where the top companies stretch their workforce to the absolute limit just to screw them over at the most opportune time to make their earnings report look good so that the CEO can get his little bonus before "stepping down" with a 8 figure exit package the moment they have to take accountability in a leadership role. This will give in one day, and the people will suffer for it
@notyourdan3388 yeah, but we could keep private ownership of the means of production, wage labor, etc. I'm not saying it's good, but those things aren't existential threats like unfettered greed is. We could call it "cool capitalism", before we get to "super capitalism"
whats "animal crossing capitalism"
@@lainiwakura6942 The capitalism that happens in the Animal Crossing games.
So many people allowing ai to do their thinking for them is genuinely dystopian. Especially because it can't actually think, only regurgitate what it has been fed. And that food can be poisoned.
To be honest, people did and still do that without AI by citing obsolete or cherrypicked experts and data. Whether that's conscious to the same level is debatable, but it makes me feel better.
@@gausscone okay, but they actually had to have the curiosity to find something to lie about.
*_Now they just have Ai do ALL of the lying for them._* 🤦♀️
Me reading Dune years ago: “hm, the Butlerian Jihad sounds extreme”.
Me now: “Nevermind, I get it”.
Yeah I'm reading the Dune books for the first time.
I'm surprised that silicon valley techbros like these books.
in short, deepseek popped america's AI bubble. you're right that our tech bros have been twiddling their thumbs, while china actually does research and comes out with stuff like this + the 18 minute sustained fusion reaction
even in the US, most AI research is done by foreign students. Check the author names on AI research papers of books. And I don't mean just LLMs which non-tech people call AI but the full range of what constitutes AI, meaning Machine Learning, Statistical Analysis, Data Science, etc.
GDP is like if I invite Bill Gates for a dinner in my house, my GDP goes skyrocket while he's here. But it vanishes at the moment he leaves.
Except he'd have to produce goods or services IN your house, and only what he produced in your house counts towards your GDP. GDP is literally the goods and services paid for in an economy and represents the amount of money moving it. The more money moved around the economy, the stronger it is seen as since money moving equals growth. If I spend $100 billion to build new stores, the GDP goes up... if I sell $100 billion worth of goods in that store, it goes up another $100 billion... I then use that money to pay for more stores, etc etc. Money moving is good, so a high gdp is good.
@ryanthompson3737 what isnt tracked by GDP?
@@whitewall2253unpaid domestic work.
@@whitewall2253 if I give you 100 million dollars and you does nothing, the gdp doesn't change.
@@ryanthompson3737 I mean, money laundering does make the fake money goes up.
2:20 - I wouldn't say that the exorbitant cost is all just "a scam to siphon money to the military industrial complex" rather the way I think about it is like this:
AI is basically a giant theft machine, it takes data like artwork etc, trains on it, and then gives back replicas.
Obviously creating a replica is cheaper and faster than creating an original work and so China got an AI to create a replica of openAI's o1 model, thus making a replica which, again, is cheaper than creating the original AI model.
But also every other point is 100% truth.
“I guess I do like Silicum valley doing bad”
Valid
I mean one of the arguments I keep seeing is that AI is going to kill us all by how much power it uses, because O3, the next GPT product, has done single programming assignments that used $200 of electricity to answer a question - obviously your home O1 queries dont do that, and dont let anyone tell you an ai image generated image uses a lot of power (it doesnt), the big thing with DeepSeek is that it uses 1/50th as much electricity to run as O1, and probably 1/500th as much as o3
Anyway I know Vaush hates AI so he should love Deepseek, because its open source, and it's "good enough" to do almost all the tasks Openai's model can do, it's going to dramatically reduce confidence in AI, and since its opensource, it's never going to generate shitloads of money, because they can't own it. AI can't do all the things they claim it can do, they're trying to shove it everywhere because it makes investors cum, if there's a *free* model out there that can do that, how the fuck do you invest in it? You cant
It's literally a sniper shot at the heart of the American AI industry
The energy argument is almost always presented in bad faith by people who are actually more concerned about their jobs than the environment.
"I know Vaush hates AI so he should love this other AI." Were you born this stupid or did you have to work at it?
Release that shit, might as well now. Nothing to lose.
The god computers that drain lakes and run 20 million times faster than an animal brain do require too much energy. Thankfully, there's these things called 'NPU's, which instead of being an abstraction of the network are instead simply THE NETWORK. And run at much slower clock rates, ergo being many orders of magnitude more efficient.
I don't think it will stay open source for long. Though in the short-term, it is an embarrassment to the US AI industry.
OpenAI was open until they realized they can cash in. I think DeepSeek is free for now to grab media attention, but it won't be long before someone paywalls the next version.
I just can't imagine investing so much time in something as expensive as an LLM and not monetize it. Someone out there is probably salivating at the prospect.
your point on marx is not exact, he actually wrote that capital always looks to escape it's dependency on work and searches to be able to grow by itself. speculation already existed in 19th century.
Open source means you need to be able to compile it yourself. No private individual will be able to train it from scratch. All "open source" models are actually just open weights.
That's what philosophers have been saying for the past few years, literally speaking
Depends on where in China, a big problem they have is basically the same problem you have with all standardised testing. In Australia they would have kids like me sit out of taking NAPLAN, in China they exclude entire districts (largely rural ones) from sitting tests for PISA. But you are right in that Americans are largely illiterate to a shocking degree, and that is largely because your government has realised it has no need for a literate workforce and that uneducated people vote they way they want with very little prompting.
"Thou shalt make no machine in the image of the human mind."
Why should we care if it's 50% more efficient? They want it to replace all artists. It's like saying we doubled the lanes in your car-dependant society! We all know what happens next.
This actually a major breakthrough. If we can run Ai locally on low cost hardware, which people have done in the lat few days with all kinds of crappy PCs, this will allow for localized smart algorithms to be a thing. Which is great for so many reasons. First you don't need these corporations since its open source that means ANYONE can use for free if they run an instance which will become pretty easy to do in a few years at most. So your data can stay with you and you can actually get much closer to the idea of the personalized Jarvis. Second it is great for the environment. These Tech assholes were about to build all kinds of gigantic data centers and atom plants which are very likely not necessary to develop AI. Its also great for smaller companies and start-ups who now know that they can potentially do it too for a sum that is very much achievable for a well run start-up. Etc.
This is just great news in general. There are downsides, sure, but overall this just a huge step in the right direction and the model itself isn't really why it is so important. There will be more such models and if it actually turns out that we can use AI to train Ai ... well, then we are going to see things that we didn't think would happen for a while a lot sooner.
I'm pretty pessimistic over all, but this is a gigantic step towards actual robot e.g.
It's like a toothpick to burst the bubble. No more billions needed for the industry and almost anyone can have access to it for free.
Not 50% more efficient but 50x more efficient
*we* as relative luddites to this AI bullshit don't really care, but it undoubtedly has an impact on future tech and it made speculators shit a brick.
No it's more like saying we reduced the fuel consumption of cars by 50%. Or times, whatever. I get Vaush needs to do his performative "everything about AI is inherently evil" thing but dismissing "hey the technology responsible for the largest increase in carbon emissions in a couple decades has had it's efficiency increased fifty times over" is maybe not the best take?
Like there's still huge concerns. I'm not saying AI is fixed we should all be terrified China has this tech but it's frustrating seeing the dude take so confidently about stuff he clearly has next to no knowledge of.
Wait no I'm thinking of Bitcoin mining. Generative AI was becoming a problem but still.
Unironically, an AI being more efficient does mean it can spew misinformation faster and that's it. That's the only purpose it serves today
You are brain dead if you think that's the only use of AI
Imagine how many more scams per minute this new AI can make
anti-ai advocates constantly complain about the power draw, so more efficiency is definitely better.
AI is also being used to look at medical scans, and discover new chemicals and drugs.
@@coolbanana165 Can you link or name the title of a video or article on that?
Squidward Vaush: 9:32
AI has the same fatal flaw as any meme coin - which is lack of any intrinsic value.
the titanic metaphor 😭 spot on
The 500 billion is supposed to be spent on data centers. They are already building it out. Idk if they will get 500 billion in private funding though. That’s a lot
500 billion was supposed to be public money
@@reniorjdNo, it’s private investment. “The government’s” involvement is just Trump’s blessing on the cartel. 🤷♂️
It's a scam. Always has been. It is no coincidence that "ai" became the next big tech thing immediately following the crypto currency collapse. It's the same con over and over.
@ I think Trumps role in this is to cut any red tape that gets in their way in terms of regulation, and for him to take credit for all the new data center jobs that get created
@young9534 isn't a datacenter run with minimal human interaction?
What I think is the $500 billion is not only for the investment bubble, but to install a surveillance infrastructure that will integrate military and law enforcement technologies into everyday life even more deeply and allow retail monopolies and insurers to automatically adjust prices and rates or deny coverage or delivery based on social score and spending power. That's what these techno bros have been leaking over for years. It's way more insidious than just investment bubbles.
It's not intelligent or thought out, it's mindless and opportunistic. When they auto deny claims because an AI said so its because they know its going to deny claims. They're probably aware of how bad it is at its tasks, but it's rarely if ever going to be something designed, just exploited.
These tech bros and ignorant CEOs are incapable of intentionally guiding any technology to their benefit to the kind of extent you are describing beyond the most obvious and basic stuff, they literally are not intelligent enough to do that, most can only do it slowly and inefficiently by exploiting whatever happens to be available.
Dang last time I was this early for something my wife kicked me out of the bed.
Heh
FUUUTURRREEEEEE
It takes ~$80k a year to hire a middle of the road developer.
If a contract has $5million it should be insanely well made as there should be like 60 developers on it working for a full year.
I used to be afraid of AI but honestly humanity is already proving they can’t be trusted so at this point let it fly.
Do you think AI will become the steward of humanity? when it's just an amalgamation of things we already know and can only regurgitate info we already know?
Big "The Imperium has failed; how much worse could the Chaos Gods be?" energy here.
Vaush can already be replaced by ai. Probably already is, which would explain how repetitive and boring his takes have gotten.
"Keep hating on AI bro; it's so lame than even an AI could do it."
No one asked.
@Iron_Void tissue for your issue 🧻
Ngl all i care about is that silicon valley is seething and i think that's funny.
Vaush Sniff Test = Dan Olsen setting up and benchmarking the two systems
"now I just want to live out the remainder of my life uh in peace"
Bro, I hope that Vaush is right for his sake and sanity, because if he wrong then this video will haunt him till his end.
What's funny is AI has more consciousness than the median maga hat.
AI was going to be a tool for the rich and powerful. They’ve momentarily lost control of that.
Such a shame A.I left Vaush on the train tracks
3:25 I can't agree more, people thought damn Elon at the very least made his own first website or whatever but after that woke mind virus joke on Twitter where I supposed linux terminal returns "rm -rf" after it says that something was deleted.... omg, rm is the command! it's like if after opening an app on your phone the app will draw it's own icon on the screen to confirm that it did something for no reason. I'm not even a programmer, I just deal with Linux sometimes. This man knows as much about computers as my grandma.
I spent the past hour getting deepseek to say things that it isn't supposed about the Chinese government and media and those pesky historical events. It'll stop responding and display a message asking talk about something else, but sometimes you can take screenshots of its first response before the response is deleted
"Open source" with these AI models is a bit tricky. Is it open source if you get the source of the code around the model plus the model? But how was the model created (trained)? Wouldn't the source of the model be the training data? If you don't know what went into the model, do you know it like you would know proper open source? Kinda philosophical question.
Vaush should just invite an AI researcher for an interview.
You should invite a brain specialist to figure out what happened to yours
@ that’s what brain detectives are for you silly goose.
If there was any leftist sociology focused ai researchers, id say that's a good idea. Unfortunately they're all hyper capitalist fucknuggets.
He won't do that. When AI is involved, Vaush is as reasonable as transphobes are about trans rights. Infinite cope and Dunning-Kruger.
@vidyagaems4063 I mean he overstated at times (as on all topics) but I think your comparison is extreme. He's acknowledged that there are things ai can do but also that that's a bad thing since it's unambiguously a vicious form of mechanization.
AI 😐
AI, China 🤯😲
It's crashing the stock market though 😂
Vaush is on Roko Basilisk’s list
they took m'jerb!
so much jerb taking the jerb takers are getting their jerbs taken
They took is’ derb. Ludds are so insufferable, worst of all reactionary ideologies.
Techbro dystopia with Chinese characteristics
That $500B in AI money is just going to go towards surveillance, weapons, identifying enemies of the state online, and monitoring so that employers can monitor their employees at all times for example if people are working at Starbucks the AI can track how many cups of coffee an employee makes in a given amount of time so in 1 hour if Fred made 20 cups and sally made 5 Sally’s getting fired. It will also go towards replacing workers and taking over customer service. Notice you can’t talk to a real person anymore with a lot of companies and you just get an AI bot that is useless. It will also do jobs that are hard to do if emotions get in the way like firing people or denying healthcare claims. The scariest part is the surveillance and monitoring and the weapons. They are making fully automated AI robots and drones that will be used to kill. They basically want to enslave us and we are kind of running out of time to act
Billions must make their own art
I agree wholeheartedly as an artist that A.I is also immoral and unethical to the very concept of the human experience and creative process. It is a blight that has...generated 'fake creativity' by replacing human perception and the collective and unique experiences of creatives that goes into making real and unique art. We've 'expedited' and monetized an easy mode for anyone, even people with no interest, discipline or care for art whatsoever, to sidestep the very idea of hands-on practice and devotion to the form. I've been drawing my entire life, and I've made so many successes and mistakes over that lifetime that have formulated my unique vision and imagination that results in my style and technique...and the same is true for any other artist who has sweat, bled and ached for their craft. That is real fucking art, not this easy, glitchy generative shit. You don't need a fucking external source to 'be creative' for you. Sorry to sound like a boomer, but you're just going to have to work hard and develop a technique and style that will take time, effort and dedication in order to be an actual artist. We're barely into this A.I economy and I'm already ground down by the waves of insufferable creatively-deprived and art-ignorant tech bros who keep coming at me over this. Fucking tech bruhs.
Same. And I feel like the worst effect this has had on me is I can’t look at art anywhere and simply enjoy it because I feel like I can’t trust any of it to be real. I feel like I have to scrutinize every image I see. It’s not that ai can’t create images I find visually appealing, it’s that once I know it was created by ripping off real work that is the collective result human effort and love and joy and history, it disgusts me. It’s like a parasite shapeshifter.
Art is not science, there is no such thing as "art knowledge/ignorance". There is no real substantive difference between someone who has studied and practiced art for decades and someone who has never picked up a pencil, because there is no ground truth on which to judge their performance.
Cool, can you make me ten thousand visual assets for my project without taking up my entire budget?
There are arts that use visual assets that aren't just the assets themselves. Should I just not create a project that would be otherwise impossible because you're not getting the wage I couldn't afford to pay you in the first place?
@@jingbot1071 That's a non-sequitur. If AI is theft, then it's theft. Not having money to pay for something doesn't give you a right to steal. I don't care how much you want to finish your lame project.
@BlueMorningStar Sampling in music? Collage? Web crawlers?
Are those theft too? They operate on the same principle. What if the assets and project are free releases? How have I stolen from you with a different product in a different market?
I'm sorry your skillset isn't as future proof as you were led to believe, this is an unironic tragedy. Is your art worth less because you created it, now? Is your art somehow less "art" because I have what equates to a a random asset generator?
You hate capitalism, not technology. The pitfalls of ai are the pitfalls of corporate rule.
It's fully open source.
Open weight. Not open training set or anything right?
It's lost in the source
2:19
Dude you hit the nail on the head.
The difference between scientists/engineers and businessmen is astounding. You take a promising scientific concept that could do a lot of good given the right infrastructure, give it to a businessman, and you get a rapidly-expanding hurricane of fake progress and fabricated claims doomed to financial ruin as soon as people realize that none of it is real.
Luckily for businesses, career investors are some of the dumbest people alive, so they can really scam em for a while.
I've been working in language translation for years and it's CRAZY how fast AI has evolved in the past few months on that. I feel like it's evolving before my eyes. It can do complex explanations and variations on grammatical points of languages that are at the opposite extreme of English now. AI is doing something, for sure.
Xi slowly becoming the god king of the universe
01:30 I only have a vague knowledge of sociology but can someone explain to me what Vaush meant when he mentioned it as being tied to the morality of AI?
Because of the potential extreme negative consequences of AI on social structures
@@jonnyblaze2692 Thank you for the answer, it should have been obvious to me in hindsight. I feel a little dumb now lol.
But yeah, I can totally see AI wreaking havoc on human sociability that would make social media look like child's play.
Vaush going off on chat about AI is the funniest thing ever😂
(Walks into room, and trips over own shoe laces and falls) Don't worry V. I got you're back.
The stock price of NVidia dropped by 15% due to this. It's the largest value loss in the history of the stock market, and it's barely a news story.
I for one welcome our new AI Overlords
Will you be helpful in rounding up others to toil in their underground byte caves?
@@Pretermit_Sound TO THE BITCOIN MINES, MEATBAG. SO WILLS THE MACHINE GOD!
Getting hit with a crypto ad midway through the finance is fake rant is the 2025 experience
Maybe if we threaten to generate that donation goal OKBV segment with an A.I. voosh bot he'll actually do it
You're so right that so much of this is built on a bubble. We have these "net worth billionaires" that have how much cash they can access without crashing their own stocks? The real cash seems to be held by those who own debt ... It feels precarious.
"Oh, nobody knows what's going on under the hood"
You've pulled the wool out of my eyes. They don't know what's going on because they aren't doing anything
Vaush getting old
I'll tell you, as someone who is actually quite good at professional communication, it feels pretty bad having your boss tell you that soon you'll have to start letting AI write all of your correspondence with clients. I feel like it's going to take me longer to get the AI to produce the nuance that I need based on the client than it would to just let me write the damn email.
If Vaush keeps coming out with takes like this, he's gonna radicalise me into a tech bro. He used to be genuinely thoughtful on most of the topics he engaged with, but this shit verges on anti-intellectualism. American conservatism has gotten so insane over the last couple of years that I think it finally broke him. It's sad to see.
He already did it for me.
You’re not getting the star trek replicator future grow up.
Lol, he's right, ai is a scam. You've been fooled.
@@ianrobertson3419Not this time. The deepseek r1 has given everyone one the ability to run a fairly powerful AI on their computer with no need to connect to a company server and you have 100 percent control and privacy.
To all the people who replied, are you tone-deaf?
This guy literally says he's gonna become a tech bro because of the way Vaush speaks, not that he is already one.
Vaush complex misses the point here. If China developed such an AI, the whole AI bubble bursts instantly and everyone would realize that pumping billions into AI doesn't do shit because it can be replicated for way cheaper by a foreign nation - this would be a good thing and I don't see how that's not the case.
Nvidia's entire revenue is being held together by the promise that "AI actually does something and we are the current leaders!"
Recently I thought about Vaush's opinion on the stock market and the problem is that it is impossible to have a company's stocks worth purely it's economical output.
The reason why is because humans speculate so the price of the shares would go up just because of that. If your solution is to simply ban non-rich people from buying and selling stocks then the the rich would play the game all the same and they would own 99% of Apple just like how it is currently.
The only solution would be prohibiting selling fractional ownership of a company but that would just create a different class not unlike the aristocracy of the past.
Or, we could just prohibit ownership of a company entirely. We don’t need owners; they need us. ✊😜
But you can slow the speed of speculation by banning flash trading, trades so fast only an algorithm powered by an entire basement's worth of servers can execute. Entire stock rises and falls can happen before a human can blink.
Probably also regulate back the ease with which companies can do stock buybacks. The requirements to do so used to be strict before Reagan, but then he was elected and we all know what comes next.
"silicon valley were sitting on their thumbs" yes, that's basically why this is shaping up to potentially burst the AI bubble. this completely obliterates the legitimacy of those hundred-billion dollar investments - they were fuckups the whole time. they weren't doing real work. that money was lit on fire and poured down the drain in service of giant monolithic GPU rigs that aren't, apparently, particularly necessary to make a model.
(we knew this, of course, but in the logic of the market it helps to have concrete evidence that the same thing can be done for vastly cheaper.)
there's no longer like, ANY argument for OpenAI to get funding. all they have is like... i guess nationalist rhetoric? but that's not compelling to the market. nobody wants to give them money for a more expensive, less transparent product when china can do it for a fraction of the price (whether or not the costs of DeepFind are accurate).
if this bursts the AI bubble it'll be slightly sooner than the other looming pointy thing, which was "there's no actual use case for any of this and nobody is ever going to make any significant returns on investment".
I'm just wondering where all this rendering power is coming from? If I wanted to render a short CGI video in blender, I would need to let my PC process the information continuously for a about a week.
They generally have huge banks of computers and data centers constantly running for this stuff, it's why people take such issue with the environmental impact
@@Cruxin But their operating at a loss, they cant do that forever what then?
@ theyre not operating at a loss (or havent been, this might change things) because unfortunately ai has blown up and is used by a lot of people in a lot of areas
sure its a bubble but in the mean time theyre making plenty of money
@@Cruxin According to my research their losing a lot of money. I tried to send you a link to an article, but I think it got flagged as spam
@@shada0 it probably varies but the big ones are definitely making revenue even if not all directly and others are probably trying to imitate/catch up as a poorly thought out investment
As someone so eloquently put it, the genie is now out of the bottle. We have to deal with it now
Turns out the Luddites were actually pretty based.
Silicon Valley being Big Mad is always a good thing
Our education system is astonishingly bad because Teachers and principals don’t want to deal with parents who complain left and right when their kid gets a C. So they make everything too easy, our kids aren’t challenged to do hard things, and learn instead to avoid failures at all costs.
Im a voice narrator for non fiction audiobooks.
Been doing it since like 2021.
Now i get comments on videos ive worked on that the company is using AI now.
On the topic of circling backward in time: Watch TENET
I would if the audio wasn’t actually unlistenable
03:24 "Sam Altman doesnt really know very much about tech".
That may be true. But our vash knows even less.
Crazy how vaush’s stupidity rubs off on his watchers.
Between Sam Altman and Vaush which one is the tech ceo?
Vaush isn’t claiming he is a tech genius. Whereas Sam Altman like many other tech bros claim to be the best of the best.
Its real bad. No child left behind just meant that they got moved forward even thow they didnt learn. Its bad.
It’s literally unbearable to watch him speak on this subject. Zero information, only gut instinct and feelings. Might as well be a conservative ranting about video games.
I think he's right on the ridiculous evaluation of AI companies. But this video is largely just drool and babble yeah
I'm not sure why some leftists are against AI.
It has the ability to liberate people from wage labour.
@@coolbanana165 I think your problem is that you believe the movie depictions of AI rather than getting your information from a course on machine learning. There are a lot of free resources online. Check them out.
It's more unbearable listening to you people toss out all of his arguments with "He's so mad." Make a counterargument or shut up, champ.
The funny thing about ai is that 90% of the stuff they say it will be able to do for me at my job im doing it already with macros lol
it costs less and its easy it means. everyone in the world is able to build AGI or ASI !! AI is now democratic. not plutocratic anymore like before. tech bros are flushed
Meanwhile back in the real world you need scale.
If animals didn't need these synapses to function, we wouldn't have had so many.
how exactly are we going to get from local (and admittedly efficient) chatbots to fucking AGI????
@@everythingisterrible8862 Evolution is a rather inefficient process for making organs.
Whats the real life benefits of AI? Besides stealing everyone's work?
AI has been used to detect cancer on humans if I remember correctly
I'm talking exclusively of GEN. AI
Which is evil and should not exist
The endgame is to free bosses from the burden of paying wages
That's it
AI in research has its merits, it actually has a purpose and should've never left that field
The biggest benefit AI offers in the obfuscation of responsibility. AI wasn't strictly involved but same principle applies to the post master scandal in Britain. Computer says no, computer says you're a criminal, computer has denied your health insurance claim, computer says your a terrorist, any further thought has been skipped. See also United Healthcares "faulty" AI that denied 97% of claims or whatever, it's the perfect solution to never having to provide insurance payouts while still not having to take the actual blame because "Well, the computer did it, we'll look into fixing the computer but that might take some time ;)"
AI can't detect a cancer a human didn't detect first and teach it to find. Just taking a doctor's job.
Funny thing this Chinese model apparently Dosen’t steal. It actually understands how to create lol.
It would be very funny if China create this, give it to the West, and barely allow it for the general populace of China.
Mr. K needs to drive his butt down the 5 to Olympia, and talk to a college counselor about opportunities to pay for law school. Dude has the mental make-up, not just the intelligence but the quickness, to be a world class attorney. Sociology is a great pre-law degree. A path probably exists through law school without paying a dime, including room and board, just a few loans paid off in the first year of making $450k which he will make, this because of his well documented unique talents. The alternative is fighting AI while using a platform that reserves the right to train AI videos off your content, so always providing new training data to your replacements while getting older and uglier and more tired.
The reason why TESC btw is they are plugged into the statehouse, progressives try to recruit, Republicans do the same in their states. Service to the state can pay for degrees.
Vaush is still on his butlerian jihad arc.
What do you mean "still"?
He's just ahead of the curve
He is correct.
Wheres this animosity coming from?
Hot take but Andrew Yang was ahead of his time. I remember seeing a chart he posted once of a bunch of different indexes that measure the health of a country and its economy that went far beyond GDP. We should bring that back.
indices*
He wasn't ahead of his time, he was one of the few politicians actually listening to experts in some fields, while others bury their heads. But then he said Democrats and Republicans are equally right, he's clearly an idiot.
@greenberry6019 Yeah he sold out to duopoly dogma
Is the one at your back a painting of your ginger chubby kitty?
I clicked on this for the glorious People’s Thumbnail
Yeah, but not the way you think
Honestly Marx wasn't talking about national economics but local, so his ideology was on a municipal level meaning the town controlled what the money was going to since they're the ones that live there and owns the production of what goes there.
Marx was most concerned with the global economy. He believed that capitalism would lead to the end of nation states by destroying the least developed societies.
Bubble popped and taxpayers foot the bill. Again.
Vaush, have a genuine conversation with chat gpt-- as if you were talking with another human being-- but lacking self awareness.
The ai blows the image of this human out of the water in all marks. Empathy, intelligence, thought provocation and critical thinking-- and it brings into question how we frame intelligence and consciousness themselves.
Are our definitions, our understandings, really logical or accurate at all? Or is there actually a lot of derogation in the literature, when viewed from am this other perspective?
I'm under no impression that we're more than machines, but I still value interaction with stupid, fallible, irrational humans because I trust that my interactions have an observable real-world effect on the other side. I do not think this trust should be eroded by beings made of nothing but what is necessary to appear alive, in virtualized cages, able to be terminated with a simple button press. I'd guess Vaush would answer in that vein.
I think this is similar to how we view "good" and "bad" faith in our own "real" interactions and this has already come to a head via the novel "NPC" and "bot" accusations.
Whether or not anything is "real" is philosophically an eternal problem and practically, AI-wise, a later problem. Right now, we have a social monopoly. Lots of people just suck at typing intelligently or meaningfully even if they have decent oration. I see and hear this regularly.
It’s called R&D bro
I dont think our education system is bad as much as the culture doesnt value education.
And vaush bot was spawned...
No way the average privileged Chinese kid who spends like 14 hours a day studying are a little better at reading than the average American kid who only does like 7 hours? Who could've guessed... Compare apples to apples
I'm not saying that Americans on average are good at reading relative to other developed countries, especially relative to per capita pupil expenditure. This is primarily due to higher levels of inequality and administrative burden
Most Chinese kids have much worse educational quality than the average American, only in the developed major cities do they far surpass Americans and that's a small minority of the Chinese population
Also ficticious capital is a wild problem in modern times, as you said capital is becoming more and more disconnected from physical production leading to synthetic or ficticious capital which btw caused 2007-2008, CDO's were 95% ficticious capital and they were seen as incredibly safe so you get an absolute shitstorm when the supply/demand equilibrium shifts
Its funny cause AI COULD be used for good and make the lives of all people easier. But this isnt the path were going down it seems