@BuildsFutures 0 seconds ago WE DONT CARE ABOUT BIG COMPANIES BENEFITNG FROM AI (And we never will) HOLY CRAP WHY DOES EVERYONE THINK AI WILL FAIL JUST BECAUSE BIG COMPANIES ARE LOSING MONEY OR CANT ADOPT IT. Its all about the small teams of 1-8 people. thats who benefits. People forget that we can DOWNLOAD QUALITY LLMS and STILL RUN AGENTS so if AI fails SMALL TEAMS WILL DOMINATE!. Big orgs are failing because of bureaucracy and big companies using methods in where they aren't AI first and really building around the Tech. I am SICK of the BIG COMPANY NARRATIVES!
Don’t get me wrong I thoroughly enjoy a lot of his content but there’s a lot of irony in him talking about how no one is using AI any more while 100% using AI to streamline his content generation flow. On a personal level my AI adoption rate is increasing significantly MoM. Probably averaging 45-90 minutes a day right now.
@rickm9244 The not-so-hidden secret of a for profit "healthcare" sector is that their incentive is NOT to cure anyone. A patient cured is a customer who can no longer be fleeced for money. It's much more profitable to make something that can be sold to people forever to keep them alive. The ultimate money maker is a drug that has a completely captive audience that will literally die if they don't get it, but will never be rid of the underlying disease. Yet braindead idologues think that private healthcare will lead to utopia because "hurr durr muh competition".
This is not the first comment I’ve seen making this reference. And as long as slip exists, I hope to see it again. May this weapon serve its purpose, and retire, when the job is done.
We wanted AI to better predict the weather, cure cancer, unravel cold fusion, etc. We don’t want to waste a city’s worth of electricity just to make infinite porn and commercials that instantly get booed.
AI has been helping weather prediction for decades, it’s not possible to cure cancer anymore than it’s possible to “cure bacteria”, but cancer detection and treatments have been using AI for decades, cold fusion is a scam, but AI is how plasma is intended to be controlled on future tokamaks. It’s doing what we want, and has been, quietly, for decades, but people mostly didn’t talk about AI till LLMs came along
@HALLish-jl5mo Yeah that's the point. They appropriated the name of one of the most promising technologies in the last decade and completely ruined public opinion and investor's confidence against it with wasteful nonsense while insulting everyone's job, stealing people's IP, making our entire technological infrastructure more dangerous and less reliable... it's insanity.
@Regicide-DSwe should just start being specific, whenever I talk about AI, I specifically say "LLMs as AI". Because LLMs are not intelligent and cannot reach AGI (artificial General Intelligence), nor do I think we can overcome their inherent flaws like hallucinations.
Even if they wanted to, the AI companies signed a contract to purchase this RAM, effectively purchasing up production capacity until 2027. Those contracts will be fulfilled and they will be paid out. If needed through the bankrupcy of those AI companies. They cant just go "nah nevermind on the RAM"
@GiegueX Yeah but that's bad. You do see that that's bad, right? It doesn't matter if lots of losers and suckers are using the brainrot algorithm, it's not nearly enough for these useless programs to make money, and that's good.
@GiegueX I use LLMs at work a lot, but the longer we have used it, the more we've found that it does not increase our productivity all that much. When it works well, it is a HUGE improvement in productivity... unfortunately, when you look at it as a whole, it wastes so much time trying to prompt for things that can be done more reliably by hand, that it has not actually improved our productivity as a whole. Right now they are so cheap to use that there is little reason to cut back how much we use it, and there is little risk in trying to use it for things that may or may not be useful, but once they start charging us enough for LLMs to make a profit, we will immediately have to stop using them for 90% of what we use it for, and the LLM customer base will tank. Bosses will crack down on you using your LLM to make a 3 sentence emails, or using their work accounts for personal use. Developers will stop asking it to add a 20 pixel margin to a button because they are too lazy to open the CSS file themselves. Graphic designers will have to start using AI generated stock photos instead of generating their own AI images because it will be faster, cheaper, and higher quality when the cost is split between users who need similar things.
yeah, you'll automatically be labeled as Scammer & Fraud if you use that. if you use it for your business, you'll lose reputation, and nobody want to touch anything your company produced
@LogicallyAnsweredhead deamon dreaming of squeezing krishna another season best believe in pnut peeling eating as i reel in fish on a new dish ❤❤ HARRRRI KRISHNAAAAAA ❤❤
First they trained us to have a fishs attention span and then, when we lose interest on AI generated slob, they complain we have a fishs attention span. Never satisfied!
See, they wanted to shorten our attention span so we forget to stop paying them when we leave their service. Not so we stop using their service AND stop paying them
They’ll probably sell it for super cheap if they close down. Then maybe people can use the data centers for something cool that doesn’t use so much water
@isaacalberda250 well with how things have been going in the last 6 years now, they will probably sell it for "cheap" but seeing as RAM has increased in price by like 4x the amount it was before, "cheap" will be something like 2x the price it was originally, and RAM prices will never go back to how they were
@Echo-hg7gu if the market is flooded with RAM initially intended for AI data centres that have not gone into operation, the price will come tumbling down. Basic supply and demand...
@Vid_Master Not really their problem is it. The value of labour just decreases and new types of jobs will have to form. Same as with the industrial revolution.
@MsFallenPrime This is not the industrial revolution now is it? During that time yes jobs were being removed however new jobs were being created. This is not like that. This is just removing jobs to replace with AI it is a net negative on jobs driving unemployment down and pushing more people into poverty when already in the US over half that country is already living paycheck to paycheck. It is ridiculous to think otherwise or even defend these corpo scum who only want to make more and more money at the expense of the people.
Fr, what were they thinking 😂 they marketed it wrong since the beginning, if only they marketed it as productivity boost and not as replacement it should be way more accepted
@Immi1897 I had to look it up and.......... * inhales * AAAAAAAAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAH AHAHAHAHAHAHAHAHAHA ahahahah ha * coughs * * looks at it again * AAAHAHAHAHAHAHAHAHAHAHAHAHAHA * cough * AHAHAHAHAHA
What i think doomed AI is how these ceo's kept trying to make ai "replace" a human. Ai does not work in a human position. Ai can't draw, can't hear music and so can't make music, can't FEEL true emotions and so can't write a genuine poem. Ai CAN'T replace a human Ai only works in areas were it can ACTUALLY be of assistance to people but even those cases are rare
Fr, if they changed their mindset to be humans living in co-existence with AI, AI being the tool to better our lives, things might be different. Right now they just hate us
I always knew they were gonna push it in this direction, Ai assistants were just the beggining. AI is the ultimate dream of a CEO, a machine that does the work of a human without the need of worker rights, not even a salary. Of course these Companies are fighting tooth and nail to make us dependant on it, this was their plan the whole time, to profit. They never intended to use it to cure cancer, they just want workers that work 24/7 without any kind of compensation, robots, basically slaves, i think that's even what the word robot means, a slave. But i really wonder what they are gonna do when everyone is too poor to afford their services and products, thanks to them we can't even have RAM for our computers now and the internet isn't useful anymore because it's filled to the brim with AI slop, they are legitimely pushing people OUT of the internet in their quest for this dream robot worker of theirs.
exactly. Ive been saying this directly to ai bros that theyre actively MAKING people hate Ai. When you're selling to PEOPLE, this doesnt bode well. I tell them if they truely believe in Ai, they should also support regulation so that PEOPLE won't be affected negatively by it & maybe WANT to use it. Instead its just inflating ram prices, destroying the environment, generating c0rn of kids, all while actively threatening to replace the very people theyre trying to sell to. Are we fr? The Ai bro's usually have a weak defense for why they don't support regulation or get extremely pissy. They seem to believe in Ai and want others to aswell , but when given the recipe to do that, they cross their arms and call us luddites.
"Ai can't draw, can't hear music and so can't make music, can't FEEL true emotions and so can't write a genuine poem. " I think this approach is dishonest and pointless. You are adding subjective categories to prevent AI from being "capable". The AI can be great at something, it can be perfect at something, yet you can always say "yeah....but...it's not doing it with EMOTION." Like sure, that's true, but if the product is the same, the emotion behind it is useless. If I want a picture made, I don't care about the "artist's process" behind it, I just want the picture. There are some things where the concept of "the human touch" is important enough to sustain the industry, like with singers and poets, but for the most part, putting "emotion" into your work isn't necessary.
Guys Remember Sam promised a cancer cure and all he did was secretly delete the millitary clause to cause more deaths. These guys are not your friend, the same "you wouldn't steal a car" people have stolen your text, videos and music
Not even in another year if you cant afford one now your waiting at least 3 to 5 years cause it bursting wont change anything for a while ram prices wont go back to the original at most you'll still pay 50% more then before the ai boom maybe 70%
Why are people not using a tool they didn't ask for? Why are people complaining about the tool they didn't ask for making everything more expensive and worse? Won't the common people please think of the shareholders?!
@andysmith8031 mate, gasoline cars are also subsidized to hell, especially fords F-150's. The difference is that electric cars are getting vastly cheaper, batteries vastly better and charging faster and cheaper and can be charged via solar. Gasoline cars still just burn their fuel and doesn't recharge it. Electric cars is the future if you're not American.
I'll stop just long enough to help them dig deep enough that my rope can't go any further, then I'll climb out and pull my rope up so they can continue.
We warned these billionaires that nobody's gonna use AI and that the bubble was going to pop. What dud they do? They ignored us and are now suffering from their idiocy.
Give it time. The real customer for AI is enterprise. Companies are willing to pay a massive amount to have AI absorb some of their labor needs. If AI can do 25% of the tasks that their employees used to do, then the company can redistribute work and eliminate up to 25% of its workforce. If they are saving tens of millions of dollars on labor, they will be willing to pay millions to the AI companies to use their products. I'm not excited for this, but I don't see any way of avoiding it. We're all gonna be like the people in wall-e in the future.
They keep telling us that AI will replace a majority of (our) jobs in the near future. I don't think they know that's not what we want to hear. They are literally threatening us to our faces.
It's because White collar jobs are already being replaced and is expected to keep increasing. That's why you see it. They are doing PR not because, "The Industry is cratering" but because its a good way not to be held accountable if they are literally telling you people are going to lose jobs and urge the government to either regulate or have plan ready for when it happens. Btw, just because some guy on RUclips tells you A.I. is going downhill, using only one company, doesn't mean its true. Research or watch videos on companies, video game consoles, ect, and you'll see the first company in line to offer a gaming console gets a ton of people buy their units and they stay on top for years until other gaming consoles come out. What happens? A new leader eventually takes its place. OpenAI is losing users to Gemini and Anthropic and the reports show massive growth for Gemini Year over year, MORE than what OpenAI is losing. That is adoption increasing, not a bubble bursting.
Well, I can tell you at least with Sora 2, it was fun at first but when you realized it blocked half the things you actually wanted to generate, it became very lame and I think dropped off real fast. It constantly false flags my generations, it doesn't allow random things like apparently you can't even make someone fatter because apparently that's offensive so it blocks doing that. As someone with a group of friends who likes to just joke around, the AI blocks 90% of fun jokes you could pull on your friends. And it blocks all copyright stuff or even describing it at all even generically, including even simply using lyrics that are famous. Doesn't take much to figure out why people aren't using Sora 2 as much anymore.
It's wild how one of the most subsided industries in human history is failing, in America in particular the tax breaks for data centers have been absurd.
And I take any future real gains from AI over short term gamer satisfaction. If we consider this "an experiment" and stop now we cannot cross tresholds.
Success according to them : Software Devs and Artists are replaced with AI. Failure according to them: AI unable to replace the software Devs and Artists
It's amazing people still don't see artists are just much cheaper than a machine that runs by itself and keeps messing up. We're already paid slave wages so it's hard to replace something that with something more expensive
@Colorstormfur Yeah I never understood why companies would just use AI instead of buying a artist who could get the job done and faster than a AI does. Cause a artist will get the specific details that a company want while a AI will just completely redraw everything if the company refresh the output.
My bank wanted me to invest in AI and I told them about my reservations as I think it's a bubble. They laughed at me. You would have thought that banks might have learned a thing or two from the last crises...
@bikesandlikes The house always wins. If a bank is about to topple, they get bailed out with tax-payer money. As long as the gravy train runs, it doesn't matter if it is a bubble. The pop of the bubble hurts you and me, not them.
Meanwhile China has none of these made up "social issues" and will continue AI research while the West lags behind being technologically regressive. Modern day luddites stopping progress in the West, which only benefits China. Gets more clearer Asia will lead and RUN the future.
It doesn't help that a large percentage of Googles AI answers are wrong. I have been shocked at how many times an incorrect answer has been at the top of my Google Search page.
I needed to get a list of Unicode circle compatibility (don't ask, Adobe Acrobat sucks is the reason), and the same number returned three different characters on three different searches (all of them wrong)
@Hotrob_J The Google AI seems to search the interwebs but does not have a context filter or even a "trusted sources" one. A funny one was that I recently watched "Wicked" with my wife and decided to ask how far the Elphiba character fell at the end of the movie (she fell for about 45 seconds). I only found one time this was mentioned on Google Search (a Reddit thread) and that was the answer confidently provided by the AI. The AI essentially quoted that one Reddit thread, which happened to be off by a factor of 5 (it did not account for wind resistance).
Yes, I've already seen that the main response from this AI can be incorrect, ranging from wrong information about a game to extremely dangerous suggestions.
The problem is, no advancement in medical tech has ever made a doctor of any kind spend more time with patients. That's the cold hard reality. Every tech advancement has made doctors spend less time interacting with patients and increased the number of patients a doctor can take on. I'm pretty sure we're all in agreement that the halcyon days of thinking technology would bring (at least America) a four-day workweek and an end of boring monotonous labor has been absolutely shattered with how the uberwealthy and out of touch tech bros have decided to employ AI and tech. Back to the doctor example. This isn't an inherently bad thing, though it tend to be used exploitatively by the medical establishment that isn't the actual care providers. A doctor who can rapidly and accurately diagnose one hundred patients a day will be better than one who does 50 a day but seems more personable. A doctor's job is to accurately cure the sick and injured. I want them to do that for as man people as possible as fast as possible with every iota of accuracy that can be gained. It's just too bad all tech people want to do is end the world for a buck.
You suggest Adobe is having some success in their implementation of AI, but the recent drama over Adobe Animate has shown that their captive customers (due to their monopoly status) are absolutely furious about AI being jammed into Adobe and their stock prices have plummeted.
Adobe has a monopoly on the professional market partly because it's really hard to switch up pipelines in industries without possibly delaying projects a few months or in some cases years. Adobe also likes to occasionally pull apple style moves and makes it very hard to change over or cancel. But there is many many many alternatives for far cheaper and honestly can be just about the same if not better than most adobe software and far more stable
@ChronicallyTired247 Well, Blender has become the industry standart pretty recently. Adobe's monopoly is not due to those reasons, it is because there is really no good alternative. Blender has had enough features to rival most modeling apps, but its UX was awful, when they fixed it Blender became a massive success. GIMP-Inkscape etc. (the whole toolset) should have, first enough features, second good UX.
They’ll take it down and claim the land so nobody can build any homes on it. Oh yeah they’ll also destroy anything inside it so they can’t be used for computer parts!
@私は神を信じますyes, but they're either blind or stupid, or they just don't care. And if history has taught us anything, it is that the most likely option is that they don't care and are hoping for a bailout if things go really wrong.
He said other companies in other sectors will benefit, not regular people. He's right, at least somewhat. Small and medium sized businesses do benefit, slowing the bubble's burst
@OnigoroshiZeroAI isn’t profitable longterm. And even then, it doesn’t have the stability or reliability to be used industry wide. It wont be for a long time. Funniling money now, isn’t wise.
@ragewireloki6213 Plus it's got a proven plateau and diminishing returns. The reason they wanna make nuclear batteries and bigger data centers for these things is because they keep hitting a wall. They're crossing their fingers hoping if they just throw more money into the hole they'll magically push it over the barrier and then they'll have a magical word/image generator that will be perfect every time, but the math doesn't support that. It was always going to fail, but they had too much momentum to listen and now they're stuck with these insane deals and obligations.
Don't be silly. It performed an extensive search on billions of sites (which, of course, tripled your electricity bill) and determined that you wanted food to be cooled. It worked great! 😀
I feel like this somewhat overlooks the fact that adobe is almost universally hated by the people who use it. They didn't crack the code, people just have to use it due to established pipelines.
@cinifiendThey aren't squirming LOL...Maybe Sam Altman and the dude from Microsoft are..but Dario Amodei and Demis Hassabis are pretty happy right now.
I hate AI and how mainstream it has become. As someone who has had / knows people who have had creative works stolen and used to train AI so to see a this has me so happy
They didn't get caught. They did those things on camera on purpose to prove they could be trusted. In the mob to be a made man you have to kill someone with a witness. It is a way to build a trust network with dishonest people.
@YAHGOAyes smartass, you are right, they did those on camera so others in the same league can trist them, we all know how secret societies work but those thing on camera are not meant to be seen by outsiders or general population, they exist as blackmail, thats what jonathantaylor was talking about, if they do get leaked they can deny it with the help of AI.
@alexandrubaginean7369 Here's the thing you seem to be missing in your theory they did this for coverage, this has all been essentially out in the open for decades. Look into the dutroux affair and the White March, and remember also that epstein was arrested over a decade ago and more than enough files and info were available then to make connections. The sad reality is that unless a radically different counter elite can arise and destroy the current elite, nothing is gonna change. And they'd have to get through all the division propaganda and the growing state security octopus to do it. You cannot vote your way out of this state of affairs and the "people" cannot do anything themselves, because they are by nature and purpose disorganized and divided.
Sure, but us scientists now use LLMs to code CNNs, survey literature, and markup papers and presentations. Also, the best performing image classifiers are now transformers.
Even for small things like converting units from ng/DL (the one time Americans use grammes 🙄) to mmol/l when comparing assay results it saves a tonne of time... not a $10tn dollar or whatever silly number industry though. I can see training management in what AI can and cannot do, and calling it consulting, being an infuriating 6 figure salary job on the horizon.
@PeterLindstrom-x4wThat's debatable. Maximum mAP? Sure. But please tell me how to run a transformer-based model with 50 frames per second on a 2,5 W NPU for robotics. Transformers are cumbersome. Good for foundational models to train high speed CNNs.
Oh yeah i forgot about that i just wish adobe stuck with what they had and dont plan to get rid of adobe animation. I use some tools in photoshop that have ai like for erasing people in my photographs or fixing something’s but only as a tool not a replacement. Ai is meant for convenience and productivity not a replacement. I don’t use it to make art
"Adobe is a big success in AI" My bro, I can't think of ONE artist. An actual artist that is happy with Adobe in 2026 and Adobe looks to manipulative contracts and schemes to trap people into being users rather than providing a good product. Instead of fixing the mounting volume of bugs in their latest iterations they include something none of us asked for, image generation. I'm a production artist, I don't need Adobe vomiting unprintable crap in illustrator out for me. I need Illustrator to work correctly and not crash.
If it is just the illustrator, might I recommend inkscape? It is open source so no reoccurring payments and it is free of AI slop the last time I checked.
Rdnm-b7i There are a lot of alternatives some are even better or about the same they lack some features but ultimately there is a lot of replacements for a lot of adobe software out there it's just adobe software tends to be pretty standard in professional settings. Some even are a bit more stable and cost effective
Adobe is getting the same reaction as Microsoft and people are dumping their products and Windows for Apple and OpenSource alternatives. Apple let’s make a good and reliable product that people actually want to buy. Open Source communities let’s use our talent and passion to produce better software than the overpriced corporate slop that’s shoved down everyone’s throats with insane subscription and licensing practices.
User: Use AI AI: Learns from users User looses job User stops using AI that much No one to invest in the market AI: Why is the market not funding us? User: Me no have money
Capitalism at its best. Wants more profit, less workers. So go all in for whatever give more money right now. And don't think too much of everything that is not money.
@WillyKillya so should we blame communism? nothing wrong with what he said, and you being annoyed at him for not saying something more profound is weird.
CEOs: "AI will soon replace everyones's jobs!" "People will soon need to think about switching professions!" Also CEOs: "Why is people so against AI? Will you please buy our subscriptions?"
Imagine a prison director being "please don't call our mess hall food 'slop' please. Just because it looks, tastes, and sounds like it doesn't mean it is."
yeah that jumped out to me, i thought they were specifically instructed NOT to tell patients anything, but idk if thats different because english is not my first language (like maybe im thinking of the techs?)
No I don't think you're wrong, afaik a radiologist very rarely speaks to the patient and instead it's a consultant who will discuss the results with the patient. It might be different in the US and other countries, but that's how it works in the UK, at least to my knowledge.
12:10 AI does not allow radiologists to study scans "infinitely fast"... I have a friend who works as the chief architect at a company that does AI pre-analysis for radiographic images and they have doubled the output of their radiologists, but they haven't made it so the analysis is so fast its over before it begins. The hyperbole from someone so high up in these companies should be seen as a major red flag...
Look into Prima at the university of Michigan. Equal to or better than human accuracy in 10 seconds per scan, compared to a half hour. The human reviewing is already becoming the bottleneck, and it's still early days
@DakotasandsurferIdem How much of human labor is garbage that needs to be reviewed? The AI has a lower error rate than trained professionals. Both outputs would need to be reviewed in any case.
Turns out saying "Trust me bro I'll pay you back next year" only works a few years in a row when billions upon billions of dollars are on the line. Who knew that savvy investors demand returns, not promises.
On top of which, the way we use it adding damage to the whole situation. The modern work force is comprised of highly specialized, narrowly focused "skilled" workers. This was the entire reason they thought AI could replace them in the first place. But using AI as an assistive tool is unusually rapid negative consequences. Theres been a lot of referencing to a study that showed Radiologists using AI to identify suspicious signs in read outs and scans, saw a 20% decline in the ability to identify things like cancer when the tool was taken away. Regardless of how accurate/flawed the study is, it raises the question of skill atrophy with prolonged use of these tools. In cases of production, this drop might not be a problem. Because if the outcome is predictable and measurable, you can tweak or modify the system to mitigate the down sides. But in cases where novelty and identification, where the outcome is unknown, that degree of performance drop is extremely concerning. With the medical scenario, the intention is increase the SPEED and accuracy of diagnosis. But with life and death on the line, missing something tiny can have significant consequences. If the AI tool fails detection, you would still expect the human to catch it at the level they were prior to using the tool. There are SciFi setting that are wary of AI, because it is still fallible, yet utterly confident in that error. A flaw shared by many software programs, where the environment is trusted and the error tolerance not correctly defined. A world being terraformed, with a large resident colony as the main labor force building up equipment, all overseen by an AI system. The AI made a mistake in the atmosphere composition, and ended up creating a runaway ecosystem collapse. When the human overseer told the AI it was making a mistake, the AI was confidant that the human was wrong, and that the corrections the human was suggesting to it would result in an ecosystem collapse. In less than a day, most of the colony died from suffocation. The few that might have found suits or cabins with breathable air would had also died long before help could arrive. All of this represents the real underlying issue we have to figure out before wide spread AI integration can be adopted without risk. The way that we use are trying to use is to offload thinking to an external system. When computers started taking over, it was reducing the labor required for data processing. In the time since, we still suffer from the consequences of not having to do a large amount of math manually, but we've compensated for it and still retain an ability to approximate numbers in order to catch larger errors. But as it stands, we still haven't figured out how to maintain an approximate idea of what the AI is putting out, or never develop those skills in the first place. If AI is going to be do things like coding, writing, and especially decision making and data validation, we NEED the ability to quickly identify if the AI is making a mistake. In the Art and Entertainment industry, the mistakes are more noticeable, but also but of far less consequence. But its adoption into surveillance and law enforcement, places where you'd think rigorous validation would already be in place, one would think AI is inserted as a parallel work flow to information gathering. But instead, its being used to skip over the validation stage entirely, and its determinations being trusted EXPLICITLY. That is THE MOST wrong way to be using these kind of systems. If anything, its shown to have the LEAST rigorous testing before deployment.
the sky high PE ration of Nvidia indicates investors predict high growth. Even the slightest drop in the acceleration of earnings means that they’re hitting a wall
@spoon-original the thing is: any new technology within our daily market tends to explode far quicker than it did in earlier years. It may seem like its still in its infancy (which I agree with you on), but it is already far farther along than what it was at the beginning. Compare the difference of image generation from 4 years ago, 3 years ago, 2 years ago, 1 year ago and now. I'm certain you'll see a drastic difference in each.
The fact that prices for nearly everything has gone up solely because of the ai bubble is what has people pissed. Anything with chips in it went up and ALL of our electric bills have gone up so these companies can privatize profits while socializing the costs and we are tired of it.
The price is for nearly everything going up is mostly because of the inflation of the money supply. The AI bubble is a product of the inflation of the money supply
@zg-it The ai bubble is a trillion dollar circle jerk between 4 companies and because Openai has already scheduled buys for 40$ of all world wide memory for the next several years it has caused the prices of EVERYTHING that uses memory to go up since 40% just disappeared. The rest of the economy is because of greedy companies. In the US nearly every large company is making record profits while blaming anyone and anything else for why they are raising prices while an feckless bunch of morons are in office.
@IamCaleumthose greedy companies work hand in hand with your greedy government. The government prints the money. They get it first. They spend it and your prices go up before you get it.
It would be a good idea to form a company to salvage them. You could buy the data centers at bankruptcy auctions for really cheap and sell the individual GPUs to gamers and potentially turn a profit even before accounting for the scrap value of the building.
@michaelimbesi2314they're designed specifically not to support desktop computers. They don't have fans and require a server rack to have a blower for airflow.
9:48 its worth noting that Google has a simple trick that allows them to get a ton of users: they embedded their AI into their search engine, and made the AI Overview show up by default, so they can essentially report everyone who uses Google but hasnt manually turned it off yet as users
Gemini is actually useful for me, and it's a thousand times better than going to chatgpt because I HAVE the sources easily and I don't have to sign up to anything.
But you would love to have fake news, virtual interactions like this and movies and T.V. shows that represent reality. If you want to live in a real world then cast away those things.
So the novelty is wearing off. I guess they figured AI would be raising the bar and just set a new baseline that art could then jump onwards from, continuing to push into new novelties. But because this new "baseline" was achieved by something like a cheat code and basically skipping ahead, no one knows how to jump forward from it. So we get stuck, and the baseline's novelty can't carry it. It stagnates. Adding onto it, the hyper-saturation AI enables, causes it to burn through the novelty even faster
@DazieArt No more Windows, no more Nvidia. I'm happy to let big companies struggle and die. They need to be reminded that they can. Them and the "nothing ever happens" crowd who genuinely believe in "too big to fail" propaganda.
Ironically, DELL earned so many goodwill. They invest in AI but only a little. They insist not to force AI into their consumer, which is not you or me, but small to medium size corporations who replace dozens to hundreds computer every five years.
Pop that bubble and put those war criminals who destroyed water and permanently damaged nature into prison. Lock them well and throw the key into the ocean.
It's more important to redistribute their wealth and limit their influence in government, as cathartic as prison would be. Trying to lock them up would only give a few lawyers a small amount of that money.
Nah, it's more like someone invents a hammer, then someone else invents a super special hammer that is more expensive to produce, and performs worse than the normal hammer, and then keep producing this special hammer despite the complete lack of monetary return and interest from consumers.
no. hammer is usefull tool, AS(artificial stupidity) is not a tool, it is like all stupidity, only capable of creating trash and pollution that make everyone worse
The anti-ai crowd are like people seeing a hammer for the first time and thinking it's useless because they don't know what it does or how to use it. "Only dunning-kruger bros think hammers aren't a fad."
I love that for how hard they’ve tried to inject weird fake words like “prompt engineer” into our vocab, the one words that’s stuck the most is “slop”. It’s just so fitting.
The Segway was a great idea only held back by stupid laws. It worked great and would have been very useful (especially to disabled people) if most governments didn’t classify it as a motorcycle. I don’t think it belongs on this list. The engineering was good, the use case was good (though oversold), the regulatory framework outside of the company’s control was bad.
Why would people support and give money to billionares just so they can get even richer by eliminating jobs of the said people? Shocker.... The greed of these billionares is a sickneeeeessss
I can't even get sloppilot to write a functional excel function of moderate complexity (I had a coworker tell me "I bet ai could do that faster" when I showed a new excel tool that saves ~10-20 hours a month of work.... It could not even come close). It's laughable to think people try to do actual programming work with it.
@Hotrob_J You can code with Claude, but you need to make all kinds of setups to have something done and you need to KNOW what it's doing so you can fix any issue
Hahaha. New claude model literally made a 10,000 lines rust, c compiler, that was able to compile a linux distro from code. All you AI haters keep telling yourselves that.
@christiangonzalez6945 "Over two weeks and nearly 2,000 Claude Code sessions costing about $20,000 in API fees, the AI model agents reportedly produced a 100,000-line Rust-based compiler capable of building a bootable Linux 6.9 kernel on x86, ARM, and RISC-V architectures." 1 C compiler for $20k "The compiler also has clear limitations that Carlini was upfront about. It lacks a 16-bit x86 backend needed to boot Linux from real mode, so it calls out to GCC for that step. Its own assembler and linker remain buggy. Even with all optimizations enabled, it produces less-efficient code than GCC running with all optimizations disabled. And the Rust code quality, while functional, does not approach what an expert Rust programmer would produce." And shitty code
@benin2059It’s not Crypto or NFT hype when it’s literally going to be normal by 2027-2028. I keep trying to tell people it’s just becoming part of our infrastructure whether they like it or not.
What these managers failed to understand is that AI is a tool, not a worker. Buy a hammer and fire the worker who knows how to use it. You can throw it at things and it may look like you've achieved something, but in the end, whatever you threw it at will simply collapse.
Check out Cape and use code LOGICALLY33 to get 33% off your first six months ➡ www.cape.co/?
... become a shareholder
WHY IS EVERYTHING AN AD MASQUERADING AS CONTENT
Yay... more propaganda denying that AI is the 4th industrial revolution...
@BuildsFutures
0 seconds ago
WE DONT CARE ABOUT BIG COMPANIES BENEFITNG FROM AI (And we never will) HOLY CRAP WHY DOES EVERYONE THINK AI WILL FAIL JUST BECAUSE BIG COMPANIES ARE LOSING MONEY OR CANT ADOPT IT. Its all about the small teams of 1-8 people. thats who benefits. People forget that we can DOWNLOAD QUALITY LLMS and STILL RUN AGENTS so if AI fails SMALL TEAMS WILL DOMINATE!. Big orgs are failing because of bureaucracy and big companies using methods in where they aren't AI first and really building around the Tech. I am SICK of the BIG COMPANY NARRATIVES!
Don’t get me wrong I thoroughly enjoy a lot of his content but there’s a lot of irony in him talking about how no one is using AI any more while 100% using AI to streamline his content generation flow. On a personal level my AI adoption rate is increasing significantly MoM. Probably averaging 45-90 minutes a day right now.
AI detecting my cancer earlier - yes
AI helping my insurance company reject my radiology claims - f off
No money in helping people walk again or even see again. Or even helping totally paralyzed people to use their own minds to control PCs.
@rickm9244 The not-so-hidden secret of a for profit "healthcare" sector is that their incentive is NOT to cure anyone. A patient cured is a customer who can no longer be fleeced for money. It's much more profitable to make something that can be sold to people forever to keep them alive. The ultimate money maker is a drug that has a completely captive audience that will literally die if they don't get it, but will never be rid of the underlying disease.
Yet braindead idologues think that private healthcare will lead to utopia because "hurr durr muh competition".
Great point
@rickm9244 money or morals. The age-old fence we perch ourselves upon, huh?
AI misdiagnosing your curable disease for cancer because it doesn't effing work.
When customers call a product slop, it's not the customer that's the problem.
The wealthy are out of touch with the public.
It put a smile on my face knowing that Satya Nadella is bothered by my use of the term "AI slop".
Its ok but AI will win
Satya Nadella more like Slopya Nutella
@6maronashey hey hey... Don't bring Nutella into this shit show!
"Stop calling it slop"
Then nickname "microslop" was born
*macroslop
*megamassivesloppyslop
Mslp
I prefer microshaft.
why did bill gates name the company after his private parts tho
"I say let it die! Let it die! Let it die! Let it shrivel up and die!"
O'Hare but he is more convincing
This is not the first comment I’ve seen making this reference. And as long as slip exists, I hope to see it again.
May this weapon serve its purpose, and retire, when the job is done.
Un-boo
O'Hare if his morals were aligned correctly:
This but reversed.
"Stop calling our shitty product slop!"
Don't talk like that. I love watching pets driving cars and motorcycles. The trillions are well spend😃
@Leynad778 hey i think ai goon content is the future
@Leynad778Enjoy wasting gallons of water.
Ok we'll call it ai shit now
Fine, I'll be more respectful and call it "poop from a butt"
We wanted AI to better predict the weather, cure cancer, unravel cold fusion, etc. We don’t want to waste a city’s worth of electricity just to make infinite porn and commercials that instantly get booed.
And the surveillance state. Starting with ICE. Strange how once they started recruiting, the proud boys disappeared...
AI has been helping weather prediction for decades, it’s not possible to cure cancer anymore than it’s possible to “cure bacteria”, but cancer detection and treatments have been using AI for decades, cold fusion is a scam, but AI is how plasma is intended to be controlled on future tokamaks.
It’s doing what we want, and has been, quietly, for decades, but people mostly didn’t talk about AI till LLMs came along
@HALLish-jl5mo Yeah that's the point. They appropriated the name of one of the most promising technologies in the last decade and completely ruined public opinion and investor's confidence against it with wasteful nonsense while insulting everyone's job, stealing people's IP, making our entire technological infrastructure more dangerous and less reliable... it's insanity.
@Regicide-DSwe should just start being specific, whenever I talk about AI, I specifically say "LLMs as AI". Because LLMs are not intelligent and cannot reach AGI (artificial General Intelligence), nor do I think we can overcome their inherent flaws like hallucinations.
Infinite fake porn is morally better than actual porn
Can we have back all that RAM?
That depends, Are you willing to visit the landfill they dump it in when the bubble bursts? 😂
Even if they wanted to, the AI companies signed a contract to purchase this RAM, effectively purchasing up production capacity until 2027. Those contracts will be fulfilled and they will be paid out. If needed through the bankrupcy of those AI companies. They cant just go "nah nevermind on the RAM"
That's not how a massive con works.
And send some my way please. My gaming rig and my workstation rig could use some.
Server ram do not work for PC.
"Why is nobody using the lying machine that drives you crazy while destroying the planet?"
Fuck dude, it's a mystery.
Except LLMs still have a LOT of traffic. Cope.
@GiegueX Yeah but that's bad. You do see that that's bad, right?
It doesn't matter if lots of losers and suckers are using the brainrot algorithm, it's not nearly enough for these useless programs to make money, and that's good.
@GiegueXthey won’t when token costs skyrocket (:
@GiegueX I use LLMs at work a lot, but the longer we have used it, the more we've found that it does not increase our productivity all that much. When it works well, it is a HUGE improvement in productivity... unfortunately, when you look at it as a whole, it wastes so much time trying to prompt for things that can be done more reliably by hand, that it has not actually improved our productivity as a whole. Right now they are so cheap to use that there is little reason to cut back how much we use it, and there is little risk in trying to use it for things that may or may not be useful, but once they start charging us enough for LLMs to make a profit, we will immediately have to stop using them for 90% of what we use it for, and the LLM customer base will tank. Bosses will crack down on you using your LLM to make a 3 sentence emails, or using their work accounts for personal use. Developers will stop asking it to add a 20 pixel margin to a button because they are too lazy to open the CSS file themselves. Graphic designers will have to start using AI generated stock photos instead of generating their own AI images because it will be faster, cheaper, and higher quality when the cost is split between users who need similar things.
yeah, you'll automatically be labeled as Scammer & Fraud if you use that.
if you use it for your business, you'll lose reputation, and nobody want to touch anything your company produced
"We lose money on every sale, but we'll make it up in volume."
Yes
I’m intrigued by your idea and I would like to subscribe to your newsletter
They probably let their AI do the math.
@irgnedeinnamenot do the math. Predict the math
@LogicallyAnsweredhead deamon dreaming of squeezing krishna another season best believe in pnut peeling eating as i reel in fish on a new dish
❤❤ HARRRRI KRISHNAAAAAA ❤❤
First they trained us to have a fishs attention span and then, when we lose interest on AI generated slob, they complain we have a fishs attention span.
Never satisfied!
It's not about attention span, but instead is about a fundamental misunderstanding of the technology's limitations. It's about being blinded by greed.
Feudal Lord World problems
See, they wanted to shorten our attention span so we forget to stop paying them when we leave their service. Not so we stop using their service AND stop paying them
I like my internet slop quality and real lmao
wtf dude, are you 12? Just make your own choices, they are not your parents
I want my fucking RAM back.
dont we all
They’ll probably sell it for super cheap if they close down. Then maybe people can use the data centers for something cool that doesn’t use so much water
not true
@isaacalberda250 well with how things have been going in the last 6 years now, they will probably sell it for "cheap" but seeing as RAM has increased in price by like 4x the amount it was before, "cheap" will be something like 2x the price it was originally, and RAM prices will never go back to how they were
@Echo-hg7gu if the market is flooded with RAM initially intended for AI data centres that have not gone into operation, the price will come tumbling down. Basic supply and demand...
Remember when chatgpt was just a cool gimmik where you could talk and the robot would also talk
Plus, I really love the videos of people clowning the Ai's suggestions in games back in the day. Oh how things change
Costs way too much money to enable that function. It was never gonna last. GPT was always in start-up mode.
It never was.
It is still cool to use and make images, but it’s still too censored and annoying to use at times. It’s not perfect. Give it another year or two.
@LordTylerGunderson No, it is not cool to steal and chop up people's work to claim you made an image.
These same billionaires who are trying to get us all laid off and driving up the prices of everything are wondering why we don't like it.
yep, and the stated goal of AI is to do all of our jobs (with no plan whatsoever to handle the results of that)
@Vid_Master Not really their problem is it. The value of labour just decreases and new types of jobs will have to form. Same as with the industrial revolution.
@MsFallenPrimeumm, it will become EVERYBODY'S problem when unemployment goes to 35% lick that corporate boot much.
@MsFallenPrime This is not the industrial revolution now is it? During that time yes jobs were being removed however new jobs were being created. This is not like that. This is just removing jobs to replace with AI it is a net negative on jobs driving unemployment down and pushing more people into poverty when already in the US over half that country is already living paycheck to paycheck. It is ridiculous to think otherwise or even defend these corpo scum who only want to make more and more money at the expense of the people.
Fr, what were they thinking 😂 they marketed it wrong since the beginning, if only they marketed it as productivity boost and not as replacement it should be way more accepted
The 2026 Superbowl felt like the 2000 Superbowl, where most of the ads were for Dotcoms right before the bubble burst
I got major crypto/nft vibes, then the industry collapsed a year later. Remember fortune favors the brave?
I found the Jew ad the most funniest
Except worse
GoDaddy
@Immi1897 I had to look it up and..........
* inhales *
AAAAAAAAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAH
AHAHAHAHAHAHAHAHAHA
ahahahah
ha
* coughs * * looks at it again *
AAAHAHAHAHAHAHAHAHAHAHAHAHAHA * cough * AHAHAHAHAHA
Scam Altman is a Billionaire but none of his ventures have ever made $1 in profit.
Modern Capitalism!
Scam Faultman becoming a billionaire is proof capitalism doesn’t work.
That damn tribe so-called chosen by God at it again😂
He and Trump must be good friends.
*Scam Slopman
What i think doomed AI is how these ceo's kept trying to make ai "replace" a human.
Ai does not work in a human position. Ai can't draw, can't hear music and so can't make music, can't FEEL true emotions and so can't write a genuine poem. Ai CAN'T replace a human
Ai only works in areas were it can ACTUALLY be of assistance to people but even those cases are rare
That and people learning how terrible these data centers are for the environment
Fr, if they changed their mindset to be humans living in co-existence with AI, AI being the tool to better our lives, things might be different. Right now they just hate us
I always knew they were gonna push it in this direction, Ai assistants were just the beggining. AI is the ultimate dream of a CEO, a machine that does the work of a human without the need of worker rights, not even a salary. Of course these Companies are fighting tooth and nail to make us dependant on it, this was their plan the whole time, to profit. They never intended to use it to cure cancer, they just want workers that work 24/7 without any kind of compensation, robots, basically slaves, i think that's even what the word robot means, a slave. But i really wonder what they are gonna do when everyone is too poor to afford their services and products, thanks to them we can't even have RAM for our computers now and the internet isn't useful anymore because it's filled to the brim with AI slop, they are legitimely pushing people OUT of the internet in their quest for this dream robot worker of theirs.
exactly. Ive been saying this directly to ai bros that theyre actively MAKING people hate Ai. When you're selling to PEOPLE, this doesnt bode well. I tell them if they truely believe in Ai, they should also support regulation so that PEOPLE won't be affected negatively by it & maybe WANT to use it. Instead its just inflating ram prices, destroying the environment, generating c0rn of kids, all while actively threatening to replace the very people theyre trying to sell to. Are we fr? The Ai bro's usually have a weak defense for why they don't support regulation or get extremely pissy. They seem to believe in Ai and want others to aswell , but when given the recipe to do that, they cross their arms and call us luddites.
"Ai can't draw, can't hear music and so can't make music, can't FEEL true emotions and so can't write a genuine poem. "
I think this approach is dishonest and pointless. You are adding subjective categories to prevent AI from being "capable". The AI can be great at something, it can be perfect at something, yet you can always say "yeah....but...it's not doing it with EMOTION." Like sure, that's true, but if the product is the same, the emotion behind it is useless. If I want a picture made, I don't care about the "artist's process" behind it, I just want the picture. There are some things where the concept of "the human touch" is important enough to sustain the industry, like with singers and poets, but for the most part, putting "emotion" into your work isn't necessary.
Guys Remember Sam promised a cancer cure and all he did was secretly delete the millitary clause to cause more deaths.
These guys are not your friend, the same "you wouldn't steal a car" people have stolen your text, videos and music
“you wouldn’t steal a car” while they crank their massive car-stealing machine.
@kayonesoft ironically if we copy THEIR data, its a breach of "intellectual property"
I would download a car if it was possible
@mspaintclippy They were unable to stop open weights models. So we can steal too.
Text videos and music? My brother, they’re gunning for your civil liberties not your media. Way worse.
Let um burn, maybe I'll be able to afford a new computer
😂
not for another year at least, as the production lines need retooling (again)
Not even in another year if you cant afford one now your waiting at least 3 to 5 years cause it bursting wont change anything for a while ram prices wont go back to the original at most you'll still pay 50% more then before the ai boom maybe 70%
First we were waiting out the pandemic then we’re waiting out the Trump import taxes AND now we’re also waiting out this stupid bubble.
@ickyelf4549there's always some calamity around the corner, at this point, I would like a year or two where nothing much of anything happens.
The only AI we care about is in other characters' behaviors in video games.
Yea, imagine an open world games with that tech, imagine gta 6 have this
true
fore real.
100%
Yeah, none of that moderation or generative AI bullshit.
“Guys please its not slop, how could you not love the smile of your mom showing you fake cctv dog and cat videos”
- AI execs
About to be "don't you love seeing your dead relative post like the good Ole days?"
AI execs sound just like MLM cult leaders.
That's literally every venture capitalist
*LLM Cult Leaders
They really do. Lost all respect for Huang.
i read that as men love men cult leaders
Well, there's this AI cult in most of Silicon valley LLM companies..
Soo, you're not wrong.
Why are people not using a tool they didn't ask for?
Why are people complaining about the tool they didn't ask for making everything more expensive and worse?
Won't the common people please think of the shareholders?!
The average person is more likely to BE a tool than to know which tools to ask for.
"Looks like you need help. " says Clippy.
Have you guys considered they don't need to make money just stop spending money on wages
This is like electric cars. They are so heavily subsidised so people by them its unreal. If there were no insetives, people wouldnt bother.
@andysmith8031 mate, gasoline cars are also subsidized to hell, especially fords F-150's.
The difference is that electric cars are getting vastly cheaper, batteries vastly better and charging faster and cheaper and can be charged via solar.
Gasoline cars still just burn their fuel and doesn't recharge it.
Electric cars is the future if you're not American.
"Stop laughing at us when we are digging our own graves."
lol nope
Istg, lol nope is gonna be the battle cry of the 2026 revolution
“Stop holding us accountable for The List and ruining our lives!”
lol nope
Lol nope
"Never interrupt your enemy when he is making a mistake."
I would laugh if they weren't so big they're going to trash the global economy and burn the biosphere on their way
I'll stop just long enough to help them dig deep enough that my rope can't go any further, then I'll climb out and pull my rope up so they can continue.
We warned these billionaires that nobody's gonna use AI and that the bubble was going to pop. What dud they do? They ignored us and are now suffering from their idiocy.
Give it time. The real customer for AI is enterprise. Companies are willing to pay a massive amount to have AI absorb some of their labor needs. If AI can do 25% of the tasks that their employees used to do, then the company can redistribute work and eliminate up to 25% of its workforce. If they are saving tens of millions of dollars on labor, they will be willing to pay millions to the AI companies to use their products. I'm not excited for this, but I don't see any way of avoiding it. We're all gonna be like the people in wall-e in the future.
The recent wave of AI CEO’s on podcasts and the news this week desperately doing PR, tells you their industry is cratering.
They keep telling us that AI will replace a majority of (our) jobs in the near future. I don't think they know that's not what we want to hear. They are literally threatening us to our faces.
It's because White collar jobs are already being replaced and is expected to keep increasing. That's why you see it. They are doing PR not because, "The Industry is cratering" but because its a good way not to be held accountable if they are literally telling you people are going to lose jobs and urge the government to either regulate or have plan ready for when it happens.
Btw, just because some guy on RUclips tells you A.I. is going downhill, using only one company, doesn't mean its true. Research or watch videos on companies, video game consoles, ect, and you'll see the first company in line to offer a gaming console gets a ton of people buy their units and they stay on top for years until other gaming consoles come out. What happens? A new leader eventually takes its place. OpenAI is losing users to Gemini and Anthropic and the reports show massive growth for Gemini Year over year, MORE than what OpenAI is losing. That is adoption increasing, not a bubble bursting.
Well, I can tell you at least with Sora 2, it was fun at first but when you realized it blocked half the things you actually wanted to generate, it became very lame and I think dropped off real fast. It constantly false flags my generations, it doesn't allow random things like apparently you can't even make someone fatter because apparently that's offensive so it blocks doing that. As someone with a group of friends who likes to just joke around, the AI blocks 90% of fun jokes you could pull on your friends. And it blocks all copyright stuff or even describing it at all even generically, including even simply using lyrics that are famous. Doesn't take much to figure out why people aren't using Sora 2 as much anymore.
@georgemontgomery1892 i really hope this is an AI comment, because if it was written by a human you are truly stupid.
It's wild how one of the most subsided industries in human history is failing, in America in particular the tax breaks for data centers have been absurd.
8:35 I'll take good RAM/Memory prices and availability over pictures of a cat in a spacesuit 8 days of the week!
And I take any future real gains from AI over short term gamer satisfaction. If we consider this "an experiment" and stop now we cannot cross tresholds.
@ArseneGraywho’s actually gaining though…
@ArseneGrayand we will never cross those thresholds with greedy companies running the show
@ArseneGray what "future real gains" are we talking about, pal?
@ArseneGray It's not just gamers needing RAM tho, what a shortsighted argument to make...
Success according to them : Software Devs and Artists are replaced with AI.
Failure according to them: AI unable to replace the software Devs and Artists
On the Sixth Day God created Man. On the Seventh Day He realized His mistake.
or Failure to them: AI doing useful things like diagnose and cure cancer or help the mars rover function independently or predict the weather
It's amazing people still don't see artists are just much cheaper than a machine that runs by itself and keeps messing up. We're already paid slave wages so it's hard to replace something that with something more expensive
@Colorstormfur Yeah I never understood why companies would just use AI instead of buying a artist who could get the job done and faster than a AI does. Cause a artist will get the specific details that a company want while a AI will just completely redraw everything if the company refresh the output.
@TheTrueEnderkingGod agreed
AI ruined googling images. It's all AI slop when all I wanted was so see a baby peacock.
This is the most frustrating bubble. I went to college during the last crash and we all know what to look out for.
My bank wanted me to invest in AI and I told them about my reservations as I think it's a bubble. They laughed at me. You would have thought that banks might have learned a thing or two from the last crises...
@bikesandlikes The house always wins. If a bank is about to topple, they get bailed out with tax-payer money. As long as the gravy train runs, it doesn't matter if it is a bubble. The pop of the bubble hurts you and me, not them.
There is a point before the collapse where smart big money quietly exits and dumb big money doubles down. That’s where we are right now.
@ratrip How did the end of that one documentary go? "And then the banks were bailed out and everything was blamed on immigrants."
Meanwhile China has none of these made up "social issues" and will continue AI research while the West lags behind being technologically regressive. Modern day luddites stopping progress in the West, which only benefits China. Gets more clearer Asia will lead and RUN the future.
It doesn't help that a large percentage of Googles AI answers are wrong. I have been shocked at how many times an incorrect answer has been at the top of my Google Search page.
I needed to get a list of Unicode circle compatibility (don't ask, Adobe Acrobat sucks is the reason), and the same number returned three different characters on three different searches (all of them wrong)
"there are 2 ls in the word apple"
@Hotrob_J The Google AI seems to search the interwebs but does not have a context filter or even a "trusted sources" one. A funny one was that I recently watched "Wicked" with my wife and decided to ask how far the Elphiba character fell at the end of the movie (she fell for about 45 seconds). I only found one time this was mentioned on Google Search (a Reddit thread) and that was the answer confidently provided by the AI. The AI essentially quoted that one Reddit thread, which happened to be off by a factor of 5 (it did not account for wind resistance).
Yes, I've already seen that the main response from this AI can be incorrect, ranging from wrong information about a game to extremely dangerous suggestions.
It once told me the symbol of carbon was Pb.... it's C....
I've never wanted to burst a bubble myself more than this one.
bring the big needles fellas!
The problem is, no advancement in medical tech has ever made a doctor of any kind spend more time with patients. That's the cold hard reality. Every tech advancement has made doctors spend less time interacting with patients and increased the number of patients a doctor can take on. I'm pretty sure we're all in agreement that the halcyon days of thinking technology would bring (at least America) a four-day workweek and an end of boring monotonous labor has been absolutely shattered with how the uberwealthy and out of touch tech bros have decided to employ AI and tech.
Back to the doctor example. This isn't an inherently bad thing, though it tend to be used exploitatively by the medical establishment that isn't the actual care providers. A doctor who can rapidly and accurately diagnose one hundred patients a day will be better than one who does 50 a day but seems more personable. A doctor's job is to accurately cure the sick and injured. I want them to do that for as man people as possible as fast as possible with every iota of accuracy that can be gained.
It's just too bad all tech people want to do is end the world for a buck.
"Don't call it AI slop"
Keep malding, lol 😂
okay microslop lol
You suggest Adobe is having some success in their implementation of AI, but the recent drama over Adobe Animate has shown that their captive customers (due to their monopoly status) are absolutely furious about AI being jammed into Adobe and their stock prices have plummeted.
My biggest problem with this video
I'm so glad I no longer work in management, lol. Adobe was already a giant workflow disruptor before jamming AI into it.
Adobe has a monopoly on the professional market partly because it's really hard to switch up pipelines in industries without possibly delaying projects a few months or in some cases years. Adobe also likes to occasionally pull apple style moves and makes it very hard to change over or cancel. But there is many many many alternatives for far cheaper and honestly can be just about the same if not better than most adobe software and far more stable
@ChronicallyTired247 Well, Blender has become the industry standart pretty recently. Adobe's monopoly is not due to those reasons, it is because there is really no good alternative. Blender has had enough features to rival most modeling apps, but its UX was awful, when they fixed it Blender became a massive success. GIMP-Inkscape etc. (the whole toolset) should have, first enough features, second good UX.
There needs to be better adobe replacements out there.
They're trying to sell us something we don't want and then act all shocked and angry about it. xD
That formula has worked through the decades for many of their products, which is why they are shocked and angry.
For me is funny when people ask something ChatGPT and than fact check it on google, why not just use google if you need to fact check it
I want it, a lot of people do.
@brainites nobody holding stock in these companies who actually understand what is hapepning are shocked or angry lol, just wagies on yt
Oh but "we" want it.
But not that way.
That's a very big difference. :)
I just drove past the construction site for an ai data center. I wonder what will happen if the bubble pops before they even finish construction lol
It’ll make a fine Spirit Halloween location
Airsoft arena
the homeless will get their housing after all, that's what is gonna happen.
Affordable housing would be a bloody nice change for once, but we can’t trust the people in charge now can we?
They’ll take it down and claim the land so nobody can build any homes on it.
Oh yeah they’ll also destroy anything inside it so they can’t be used for computer parts!
We’re too broke and too stressed to care about AI. The audacity to think regular people would be happy with AI
Totally, like we should be so happy to pay for some chatbot that can order doordash for me
0:50 I love how he accidentally admitted it was a bubble in an attempt to deny it being a bubble.
My first reaction. Isn't only tech companies are benefiting from this Ai hype wagen ?
@私は神を信じますpretty much, the average person won't see much benefit, in fact, it's more likely to screw us hard for at least a few years or more.
@私は神を信じますyes, but they're either blind or stupid, or they just don't care. And if history has taught us anything, it is that the most likely option is that they don't care and are hoping for a bailout if things go really wrong.
He said other companies in other sectors will benefit, not regular people. He's right, at least somewhat. Small and medium sized businesses do benefit, slowing the bubble's burst
Unlike other hype cycles people actively hate AI and most professionals either abuse it for laziness or hate it for its uselessness.
The haters are the 1% of the braindead and the butthurt artists that can't cope with reality.
@OnigoroshiZeroAI isn’t profitable longterm. And even then, it doesn’t have the stability or reliability to be used industry wide. It wont be for a long time. Funniling money now, isn’t wise.
@OnigoroshiZerocry harder AI Bro
@ragewireloki6213 Plus it's got a proven plateau and diminishing returns. The reason they wanna make nuclear batteries and bigger data centers for these things is because they keep hitting a wall. They're crossing their fingers hoping if they just throw more money into the hole they'll magically push it over the barrier and then they'll have a magical word/image generator that will be perfect every time, but the math doesn't support that. It was always going to fail, but they had too much momentum to listen and now they're stuck with these insane deals and obligations.
I think that the hate is the biological hatred of machine learning turned up to 11 with how hard aibros have pushed llms so hard
Generative AI just sucks and I'd be happy if we simply banned it.
My AI refrigerator just isn't that great.
But Elon will build fridges on the moon, which will be much more profitable, because moon inhabitants have no fridges at all. It's a whole new market
@neilwilson5785 😂😂😂
Don't be silly. It performed an extensive search on billions of sites (which, of course, tripled your electricity bill) and determined that you wanted food to be cooled. It worked great! 😀
@jfess1911Nah. It searched the correct temperature for eggs, and promptly warmed everything to the body temp of a chicken...
@sacrificialrook6283 Actually....... you may have a point there.🤔
I want nothing more than to see companies like Meta go 6 feet under.
And their leaders with them
I'm fine with no longer calling AI "slop" so long as we find another, much worse word to use instead.
Ai sludge? How's that lol
How about ai bullshit
waste
Digital pollution. Oh wait, that's not an insult! That's the truth!
AI slosh????
I feel like this somewhat overlooks the fact that adobe is almost universally hated by the people who use it. They didn't crack the code, people just have to use it due to established pipelines.
12:10 radiologists do NOT spend time with patients.
Are you really surprised he has no idea how any of this work?
@Dinnye01😂
Sam Altman is the role model of the successful American: a con man, a liar, a hypocrite, and a bad person.
Sounds more like a commie
He is a woman
@degenmoves No he’s not
@degenmoveslmao
@gnosticmind no adams apple and it's body frame...you sure about that?
I’m glad open AI is failing
I'm glad we're seeing all these idiotic CEOs start to squirm now as they realize they bet the farm and it's not a winning bet.
Same
Call me immature, but same here.
Thank china and deepseek architecture
@cinifiendThey aren't squirming LOL...Maybe Sam Altman and the dude from Microsoft are..but Dario Amodei and Demis Hassabis are pretty happy right now.
I hate AI and how mainstream it has become. As someone who has had / knows people who have had creative works stolen and used to train AI so to see a this has me so happy
Pretty sure a few trillion was spent on AI just so the elites that got caught on camera doing unspeakable things can claim those videos are AI.
They didn't get caught. They did those things on camera on purpose to prove they could be trusted. In the mob to be a made man you have to kill someone with a witness. It is a way to build a trust network with dishonest people.
@YAHGOAyes smartass, you are right, they did those on camera so others in the same league can trist them, we all know how secret societies work but those thing on camera are not meant to be seen by outsiders or general population, they exist as blackmail, thats what jonathantaylor was talking about, if they do get leaked they can deny it with the help of AI.
@alexandrubaginean7369 Here's the thing you seem to be missing in your theory they did this for coverage, this has all been essentially out in the open for decades. Look into the dutroux affair and the White March, and remember also that epstein was arrested over a decade ago and more than enough files and info were available then to make connections. The sad reality is that unless a radically different counter elite can arise and destroy the current elite, nothing is gonna change. And they'd have to get through all the division propaganda and the growing state security octopus to do it. You cannot vote your way out of this state of affairs and the "people" cannot do anything themselves, because they are by nature and purpose disorganized and divided.
@satomonster568You have a point, I agree but I was talking about something different and responding to that guys post.
100%
Big, big remainder that those AIs in sciences aren't LLMs.
Sure, but us scientists now use LLMs to code CNNs, survey literature, and markup papers and presentations. Also, the best performing image classifiers are now transformers.
Even for small things like converting units from ng/DL (the one time Americans use grammes 🙄) to mmol/l when comparing assay results it saves a tonne of time... not a $10tn dollar or whatever silly number industry though. I can see training management in what AI can and cannot do, and calling it consulting, being an infuriating 6 figure salary job on the horizon.
@PeterLindstrom-x4wThat's debatable. Maximum mAP? Sure. But please tell me how to run a transformer-based model with 50 frames per second on a 2,5 W NPU for robotics.
Transformers are cumbersome. Good for foundational models to train high speed CNNs.
@juliusfucik4011 Not my field but I get your argument. I'm in medical imaging so speed is a lot less important than accuracy/ROC, etc.
Thankfully
9:26 Crazy that Adobe has an entire ad just to showcase that they can do slop, though.
Oh yeah i forgot about that i just wish adobe stuck with what they had and dont plan to get rid of adobe animation. I use some tools in photoshop that have ai like for erasing people in my photographs or fixing something’s but only as a tool not a replacement. Ai is meant for convenience and productivity not a replacement. I don’t use it to make art
The worst thing about AI is the pollution
Im surprised these idiots didn’t invest millions into fidget spinners and pogs
Alf though
And beanie babies
"Adobe is a big success in AI" My bro, I can't think of ONE artist. An actual artist that is happy with Adobe in 2026 and Adobe looks to manipulative contracts and schemes to trap people into being users rather than providing a good product. Instead of fixing the mounting volume of bugs in their latest iterations they include something none of us asked for, image generation. I'm a production artist, I don't need Adobe vomiting unprintable crap in illustrator out for me. I need Illustrator to work correctly and not crash.
Facts
If it is just the illustrator, might I recommend inkscape? It is open source so no reoccurring payments and it is free of AI slop the last time I checked.
Rdnm-b7i There are a lot of alternatives some are even better or about the same they lack some features but ultimately there is a lot of replacements for a lot of adobe software out there it's just adobe software tends to be pretty standard in professional settings. Some even are a bit more stable and cost effective
Adobe is getting the same reaction as Microsoft and people are dumping their products and Windows for Apple and OpenSource alternatives. Apple let’s make a good and reliable product that people actually want to buy. Open Source communities let’s use our talent and passion to produce better software than the overpriced corporate slop that’s shoved down everyone’s throats with insane subscription and licensing practices.
@ChronicallyTired247 I know, but it's corpo standard right now, unless I'm wrong? You're expected to use Adobe
User: Use AI
AI: Learns from users
User looses job
User stops using AI that much
No one to invest in the market
AI: Why is the market not funding us?
User: Me no have money
Capitalism at its best.
Wants more profit, less workers.
So go all in for whatever give more money right now.
And don't think too much of everything that is not money.
except its ceo falling for the hype and it really doesnt replace anyone
@_Cfocus Worse, for coding, it's like paying monkey typists per line of code.
@Thanhatos blaming "capitalism" is such a lazy pseudo-intellectual auto-response from people who can't think deeply about anything.
@WillyKillya so should we blame communism? nothing wrong with what he said, and you being annoyed at him for not saying something more profound is weird.
Seeing AI fail makes me hit the Kakegurui Masho
I feel you
It's almost like people are more worried about feeding themselves verses the new shiny tech gadget.
I wONdEr WhY
CEOs: "AI will soon replace everyones's jobs!" "People will soon need to think about switching professions!"
Also CEOs: "Why is people so against AI? Will you please buy our subscriptions?"
Imagine a prison director being "please don't call our mess hall food 'slop' please. Just because it looks, tastes, and sounds like it doesn't mean it is."
It has to be a duck then
It can't go away fast enough...
radiologists never see patients. they only look at screens. hype LOL.
Tbf, there are interventional radiologists who actually see patients. But, that doesn't seem like who Huang is talking about.
yeah that jumped out to me, i thought they were specifically instructed NOT to tell patients anything, but idk if thats different because english is not my first language (like maybe im thinking of the techs?)
No I don't think you're wrong, afaik a radiologist very rarely speaks to the patient and instead it's a consultant who will discuss the results with the patient.
It might be different in the US and other countries, but that's how it works in the UK, at least to my knowledge.
“Let make a funny content creation tool. That’ll keep them hooked”
The cooked attention span:
Imagine if someone told executives that what they were pouring money into wasn't AI, but LLM.
honestly yeah, i got bored with AI fairly quickly. I don't get how people use it as a conversation partner???
@lemin0u I roast ai bots for fun... I have tried actual conversations too,...it only makes me want to throw my phone off the window
@CatMagnet_lovescats how's that fun for longer than like 10 minutes
"Why does no one want to buy our slop!?"
You're pfp and one of your thumbnails is AI, hypocrite.
Rule number 1 of business: You don't invest in trends
12:10 AI does not allow radiologists to study scans "infinitely fast"... I have a friend who works as the chief architect at a company that does AI pre-analysis for radiographic images and they have doubled the output of their radiologists, but they haven't made it so the analysis is so fast its over before it begins. The hyperbole from someone so high up in these companies should be seen as a major red flag...
We've reached a stage of finance capitalism where not even a 100% productivity increase is enough to service the debt.
Look into Prima at the university of Michigan. Equal to or better than human accuracy in 10 seconds per scan, compared to a half hour. The human reviewing is already becoming the bottleneck, and it's still early days
@Lucifer5-f5vcomputer vision is not generative AI though.
@Lucifer5-f5vand how much of that is garbage that needs to be reviewed?
@DakotasandsurferIdem How much of human labor is garbage that needs to be reviewed? The AI has a lower error rate than trained professionals. Both outputs would need to be reviewed in any case.
Got an email today from Open Ai telling me about an update to their terms of service... ads will now be a feature
-uninstall
yep, immediately canceled my subscription. will be looking for a new LLM to help with coding.
Home labing if you have money
@saffainamerica Stop using an LLM for coding. It will only cause you to get worse at coding as you stop flexing your brain
Just use local llms with llm studio
@Vode_ikaAgreed
Turns out saying "Trust me bro I'll pay you back next year" only works a few years in a row when billions upon billions of dollars are on the line. Who knew that savvy investors demand returns, not promises.
Every sin is forgiven till line goes up.
When line goes down, everyone save themselves.
6:58 every app says that then gets exposed for selling data
0:30 “people just aren’t using ai tools like they used to” 💀 you didn’t know what ai was 7 years ago…
On top of which, the way we use it adding damage to the whole situation. The modern work force is comprised of highly specialized, narrowly focused "skilled" workers. This was the entire reason they thought AI could replace them in the first place. But using AI as an assistive tool is unusually rapid negative consequences. Theres been a lot of referencing to a study that showed Radiologists using AI to identify suspicious signs in read outs and scans, saw a 20% decline in the ability to identify things like cancer when the tool was taken away. Regardless of how accurate/flawed the study is, it raises the question of skill atrophy with prolonged use of these tools.
In cases of production, this drop might not be a problem. Because if the outcome is predictable and measurable, you can tweak or modify the system to mitigate the down sides. But in cases where novelty and identification, where the outcome is unknown, that degree of performance drop is extremely concerning. With the medical scenario, the intention is increase the SPEED and accuracy of diagnosis. But with life and death on the line, missing something tiny can have significant consequences. If the AI tool fails detection, you would still expect the human to catch it at the level they were prior to using the tool.
There are SciFi setting that are wary of AI, because it is still fallible, yet utterly confident in that error. A flaw shared by many software programs, where the environment is trusted and the error tolerance not correctly defined. A world being terraformed, with a large resident colony as the main labor force building up equipment, all overseen by an AI system. The AI made a mistake in the atmosphere composition, and ended up creating a runaway ecosystem collapse. When the human overseer told the AI it was making a mistake, the AI was confidant that the human was wrong, and that the corrections the human was suggesting to it would result in an ecosystem collapse. In less than a day, most of the colony died from suffocation. The few that might have found suits or cabins with breathable air would had also died long before help could arrive.
All of this represents the real underlying issue we have to figure out before wide spread AI integration can be adopted without risk. The way that we use are trying to use is to offload thinking to an external system. When computers started taking over, it was reducing the labor required for data processing. In the time since, we still suffer from the consequences of not having to do a large amount of math manually, but we've compensated for it and still retain an ability to approximate numbers in order to catch larger errors. But as it stands, we still haven't figured out how to maintain an approximate idea of what the AI is putting out, or never develop those skills in the first place. If AI is going to be do things like coding, writing, and especially decision making and data validation, we NEED the ability to quickly identify if the AI is making a mistake. In the Art and Entertainment industry, the mistakes are more noticeable, but also but of far less consequence.
But its adoption into surveillance and law enforcement, places where you'd think rigorous validation would already be in place, one would think AI is inserted as a parallel work flow to information gathering. But instead, its being used to skip over the validation stage entirely, and its determinations being trusted EXPLICITLY. That is THE MOST wrong way to be using these kind of systems. If anything, its shown to have the LEAST rigorous testing before deployment.
*4 years ago
the sky high PE ration of Nvidia indicates investors predict high growth. Even the slightest drop in the acceleration of earnings means that they’re hitting a wall
davexmitthat ai is still in its infancy and people are complaining about “it ain’t like it used to be”…
@spoon-original the thing is: any new technology within our daily market tends to explode far quicker than it did in earlier years. It may seem like its still in its infancy (which I agree with you on), but it is already far farther along than what it was at the beginning. Compare the difference of image generation from 4 years ago, 3 years ago, 2 years ago, 1 year ago and now. I'm certain you'll see a drastic difference in each.
The fact that prices for nearly everything has gone up solely because of the ai bubble is what has people pissed. Anything with chips in it went up and ALL of our electric bills have gone up so these companies can privatize profits while socializing the costs and we are tired of it.
This message needs to be on billboards for the world to see!!!
The price is for nearly everything going up is mostly because of the inflation of the money supply. The AI bubble is a product of the inflation of the money supply
@zg-it The ai bubble is a trillion dollar circle jerk between 4 companies and because Openai has already scheduled buys for 40$ of all world wide memory for the next several years it has caused the prices of EVERYTHING that uses memory to go up since 40% just disappeared. The rest of the economy is because of greedy companies. In the US nearly every large company is making record profits while blaming anyone and anything else for why they are raising prices while an feckless bunch of morons are in office.
@IamCaleumthose greedy companies work hand in hand with your greedy government. The government prints the money. They get it first. They spend it and your prices go up before you get it.
In less than ten years most of those data centers will be great for urban explorers, rats and pigeons.
It would be a good idea to form a company to salvage them. You could buy the data centers at bankruptcy auctions for really cheap and sell the individual GPUs to gamers and potentially turn a profit even before accounting for the scrap value of the building.
You can’t use H100 for gaming it doesn’t have a video out 😓
@michaelimbesi2314i dont think server gpus are good for gaming
@michaelimbesi2314they're designed specifically not to support desktop computers. They don't have fans and require a server rack to have a blower for airflow.
@michaelimbesi2314the hardware there ain't for gamers.
Im (Finally) Bursting...
9:48 its worth noting that Google has a simple trick that allows them to get a ton of users: they embedded their AI into their search engine, and made the AI Overview show up by default, so they can essentially report everyone who uses Google but hasnt manually turned it off yet as users
Gemini is actually useful for me, and it's a thousand times better than going to chatgpt because I HAVE the sources easily and I don't have to sign up to anything.
Tf u mean manually turning it off? If you found a way to turn that shit off then do tell!
@FogololWhen searching something up type "-ai" or "-AI" at the end of the search you are looking for and there you go
How does one turn that shid off twin, don't say you can without telling us how
@Fogolol you can add -ai to the end of your query, but i don't think there is a master switch to turn this off completely
Nobody wants AI slop
nah i love it
@debestapolishcringe are you sure?
@debestapolishcringei love it too
@GreenBean6570you're an offensive existence
@GreenBean6570it’s what it is.
I live in a 'real' World - real People, real Things, real Interactions ... I don't want to constantly look at fakes made up by some AI
But you would love to have fake news, virtual interactions like this and movies and T.V. shows that represent reality. If you want to live in a real world then cast away those things.
@RandalfElVikingoholy hell we got an ai bro that ignores all the bad sides of ai
@cat_blue_background most of the stuff that people call real is also fake.
So the novelty is wearing off.
I guess they figured AI would be raising the bar and just set a new baseline that art could then jump onwards from, continuing to push into new novelties.
But because this new "baseline" was achieved by something like a cheat code and basically skipping ahead, no one knows how to jump forward from it. So we get stuck, and the baseline's novelty can't carry it. It stagnates. Adding onto it, the hyper-saturation AI enables, causes it to burn through the novelty even faster
3:23 "Up to $100 billion"
Yeah, it turns out the phrase "up to" was doing a *ton* of heavy lifting on their bluff.
Just remember the companies that turned their backs on the ones that created them (GPU, RAM specifically).
All of em basically
I'm not getting an Nvidia GPU when I upgrade this summer. ARC has had good driver updates.
@DazieArt No more Windows, no more Nvidia. I'm happy to let big companies struggle and die. They need to be reminded that they can. Them and the "nothing ever happens" crowd who genuinely believe in "too big to fail" propaganda.
Ironically, DELL earned so many goodwill. They invest in AI but only a little. They insist not to force AI into their consumer, which is not you or me, but small to medium size corporations who replace dozens to hundreds computer every five years.
Always support amd from now. When this shit crashes DO NOT BUY NVIDIA. LET THEM DIE
Pop that bubble and put those war criminals who destroyed water and permanently damaged nature into prison. Lock them well and throw the key into the ocean.
but what about trump
It's more important to redistribute their wealth and limit their influence in government, as cathartic as prison would be. Trying to lock them up would only give a few lawyers a small amount of that money.
Generative AI is pretty bad, but you're a fanatic
@konomibaba-t8nno he just has common sense
They DEMOLSIHED the water
You never mess with the gaming community
10:56 top seven big what companies
Top seven big tech companies
😂
Probably sell a lot of Viagra to be the top seven
😮 .... 😂😂
You heard him
Open AI wants taxpayers to keep AI going so it can replace their jobs, lol
"Too big to fail." Nope. Too big to exist.
It's like somebody inventing the first hammer and then everybody expecting it to build cities on the moon in a short amount of time
More like the inventor of the hammer claiming it can build cities in the Moon by itself
Nah, it's more like someone invents a hammer, then someone else invents a super special hammer that is more expensive to produce, and performs worse than the normal hammer, and then keep producing this special hammer despite the complete lack of monetary return and interest from consumers.
I read Dr Stone. That took 3 years.
no. hammer is usefull tool, AS(artificial stupidity) is not a tool, it is like all stupidity, only capable of creating trash and pollution that make everyone worse
The anti-ai crowd are like people seeing a hammer for the first time and thinking it's useless because they don't know what it does or how to use it. "Only dunning-kruger bros think hammers aren't a fad."
Nobody wants this AI shit and you can't get away from it. You can't even Google search something with out an AI outline coming up.
I love that for how hard they’ve tried to inject weird fake words like “prompt engineer” into our vocab, the one words that’s stuck the most is “slop”. It’s just so fitting.
not prompt engineer, sloperator!
I can't wait until "AI" takes its rightful place among the 3DTVs, Segways, The Metaverse, Windows Phones, Hoverboards, Google Glass, etc.
Don't forget NFTs
Atleast Windows phone was fun. Was it good? No but atleast it was fun(had two of them)
Windows phones were peak though.
The Segway was a great idea only held back by stupid laws.
It worked great and would have been very useful (especially to disabled people) if most governments didn’t classify it as a motorcycle.
I don’t think it belongs on this list. The engineering was good, the use case was good (though oversold), the regulatory framework outside of the company’s control was bad.
@theelitelancemy sister had one, I hated it, she hated, needless to say, she didn't preserve with it for more than a few months.
Why would people support and give money to billionares just so they can get even richer by eliminating jobs of the said people? Shocker.... The greed of these billionares is a sickneeeeessss
Thank god that the AI Bubble is Bursting
12:20 That is a study using neural net machine learning and not generative AI...
Yup, it drives me insane how entire fields of study in computer science get lumped into transformer and diffusion generative models
Programming is so much more complex than people think...
These Ai won't be able to accomplish no where near what these corporations are saying..
I can't even get sloppilot to write a functional excel function of moderate complexity (I had a coworker tell me "I bet ai could do that faster" when I showed a new excel tool that saves ~10-20 hours a month of work.... It could not even come close). It's laughable to think people try to do actual programming work with it.
@Hotrob_J You can code with Claude, but you need to make all kinds of setups to have something done and you need to KNOW what it's doing so you can fix any issue
Hahaha.
New claude model literally made a 10,000 lines rust, c compiler, that was able to compile a linux distro from code.
All you AI haters keep telling yourselves that.
AI can certainly write code, but it can't engineer software.
@christiangonzalez6945 "Over two weeks and nearly 2,000 Claude Code sessions costing about $20,000 in API fees, the AI model agents reportedly produced a 100,000-line Rust-based compiler capable of building a bootable Linux 6.9 kernel on x86, ARM, and RISC-V architectures."
1 C compiler for $20k
"The compiler also has clear limitations that Carlini was upfront about. It lacks a 16-bit x86 backend needed to boot Linux from real mode, so it calls out to GCC for that step. Its own assembler and linker remain buggy. Even with all optimizations enabled, it produces less-efficient code than GCC running with all optimizations disabled. And the Rust code quality, while functional, does not approach what an expert Rust programmer would produce."
And shitty code
Bro please just a few hundred billions more and AI will be profitable bro we swear.
Won't anyone please think of the billionaires?!
no lol
Why does the hype graph look exactly like the Dunning-Kruger Graph?
The more people realizing that "A.I." is a complete scam the better.
How when 80 million+ use AI?
That's... Not how it works buddy. AI fan or not.
@benin2059It’s not Crypto or NFT hype when it’s literally going to be normal by 2027-2028. I keep trying to tell people it’s just becoming part of our infrastructure whether they like it or not.
@LordTylerGunderson That's why there are so many bullshit videos on all social medias.
Reminder OpenAI krilled a whistleblower
who?
Suchir Balaji
Boeing upvotes it
What these managers failed to understand is that AI is a tool, not a worker. Buy a hammer and fire the worker who knows how to use it. You can throw it at things and it may look like you've achieved something, but in the end, whatever you threw it at will simply collapse.
Wow
This ☝️
You can't replace a worker and his hammer with an electric hammer lol
5:09 I am not paying for ai to take my ram and gpu ✌️
Sam Altman murdered Suchir Balaji
💯
With shower?
I like how that "news" has gone nowhere since.
I heard about that, Sam is up to no good.
Boeing : Filthy Amatuers
12:30 "Deep Generative Models" are generative models used to create likely training data to train other neural networks that aren't generative...
Yeah, I think most examples he mentioned aren't about generative AI
Immediately checked the comments for this.
Dude uses AI like everything before chatgpt counts. Might not be the most reliable source to listen to
guys what hapoened to dalle mini😭