AI Expert Warns Crash Is Imminent As AI Improvements Hit Brick Wall

"LLMs such as they are, will become a commodity; price wars will keep revenue low. Given the cost of chips, profits will be elusive," Marcus predicts. "When everyone realizes this, the financial bubble may burst quickly."

Please let this happen

Market crash and third world war. What a time to be alive!

I wish just once we could have some kind of tech innovation without a bunch of douchebag techbros thinking it's going to solve all the world's problems with no side effects while they get super rich off it.

... bunch of douchebag techbros thinking it's going to solve all the world's problems with no side effects...
one doesn't imagine any of them even remotely thinks a technological panacaea is feasible.
... while they get super rich off it.
because they're only focusing on this.
- Oh they definitely exist. At a high level the bullshit is driven by malicious greed, but there are also people who are naive and ignorant and hopeful enough to hear that drivel and truly believe in it.
  Like when Microsoft shoves GPT4 into notepad.exe. Obviously a terrible terrible product from a UX/CX perspective. But also, extremely expensive for Microsoft right? They don't gain anything by stuffing their products with useless annoying features that eat expensive cloud compute like a kid eats candy. That only happens because their management people truly believe, honest to god, that this is a sound business strategy, which would only be the case if they are completely misunderstanding what GPT4 is and could be and actually think that future improvements would be so great that there is a path to mass monetization somehow.
- True, they just sell it to their investors as a panacea
- Some are just opportunists, but there are certainly true believers — either in specific technologies, or pedal-to-the-metal growth as the only rational solution to the world’s problems.
  Andreessen is pretty open about it: https://a16z.com/the-techno-optimist-manifesto/
Of course most don't actually even believe it, that's just the pitch to get that VC juice. It's basically fraud all the way down.
Soooo... Without capitalism?
- Pretty much.

No shit. This was obvious from day one. This was never AGI, and was never going to be AGI.

Institutional investors saw an opportunity to make a shit ton of money and pumped it up as if it was world changing. They'll dump it like they always do, it will crash, and they'll make billions in the process with absolutely no negative repercussions.

Turns out AI isn't real and has no fidelity.
Machine learning could be the basis of AI but is anyone even working on that when all the money is in LLMs?
- I'm not an expert, but the whole basis of LLM not actually understanding words, just the likelihood of what word comes next basically seems like it's not going to help progress it to the next level... Like to be an artificial general intelligence shouldn't it know what words are?
  I feel like this path is taking a brick and trying to fit it into a keyhole...
Then what is this I’m feeling if it’s not AGI? 🤔
- Maybe GERD?

largely based on the notion that LLMs will, with continued scaling, become artificial general intelligence

Who said that LLMs were going to become AGI? LLMs as part of an AGI system makes sense but not LLMs alone becoming AGI. Only articles and blog posts from people who didn't understand the technology were making those claims. Which helped feed the hype.

I 100% agree that we're going to see an AI market correction. It's going to take a lot of hard human work to achieve the real value of LLMs. The hype is distracting from the real valuable and interesting work.

Journalists have no clue what AI even is. Nearly every article about AI is written by somebody who couldn't tell you the difference between an LLM and an AGI, and should be dismissed as spam.
OpenAI published a paper about GPT titled "Sparks of AGI".
I don't think they really believe it but it's good to bring in VC money
- That is a very VC baiting title. But it's doesn't appear from the abstract that they're claiming that LLMs will develop to the complexity of AGI.
The call is coming from inside. Google CEO claims it will be like alien intelligence so we should just trust it to make political decisions for us bro: https://www.computing.co.uk/news/2024/ai/former-google-ceo-eric-schmidt-urges-ai-acceleration-dismisses-climate
- Do you have a non paywalled link? And is that quote in relation to LLMs specifically or AI generally?
I read a lot I guess, and I didn’t understand why they think like this. From what I see, are constant improvements in MANY areas! Language models are getting faster and more efficient. Code is getting better across the board as people use it to improve their own, contributing to the whole of code improvements and project participation and development. I feel like we really are at the beginning of a lot of better things and it’s iterative as it progresses. I feel hopeful

It's so funny how all this is only a problem within a capitalist frame of reference.

What they call "AI" is only "intelligent" within a capitalist frame of reference, too.
- I don't understand why you're being downvoted. Current "AI" based on LLM's have no capacity for understanding of the knowledge they contain (hence all the "hallucinations"), and thus possess no meaningful intelligence. To call it intelligent is purely marketing.

Thank fuck. Can we have cheaper graphics cards again please?

I'm sure a RTX 4090 is very impressive, but it's not £1800 impressive.

Just wait for the 5090 prices...
- I just don't get whey they're so desperate to cripple the low end cards.
  Like I'm sure the low RAM and speed is fine at 1080p, but my brother in Christ it is 2024. 4K displays have been standard for a decade. I'm not sure when PC gamers went from "behold thine might from thou potato boxes" to "I guess I'll play at 1080p with upscaling if I can have a nice reflection".
Sorry, crypto is back in season.
I swapped to AMD this generation and it's still expensive.
- A well researched pre-owned is the way to go. I bought a 6900xt a couple years ago for a deal.
nope, if normal gamers are already willing to pay that price, no reason for nvidia to reduce them.
There's more 4090 on steam than any AMD dedicated GPU, there's no competition
AMD will go back to the same strategy they had with the RX 580. They don't plan to release high end cards next generation. It seems they just want to pump out a higher volume of mid-tier (which is vague and subjective) while fixing hardware bugs plaguing the previous generation.
Hopefully, this means we can game on a budget while AMD is focusing primarily on marketshare.

Oh no!

Anyway...

I've been hearing about the imminent crash for the last two years. New money keeps getting injected into the system. The bubble can't deflate while both the public and private sector have an unlimited lung capacity to keep puffing into it. FFS, bitcoin is on a tear right now, just because Trump won the election.
This bullshit isn't going away. Its only going to get forced down our throats harder and harder, until we swallow or choke on it.
- With the right level of Government support, bubbles can seemingly go on for literal decades. Case in point, Australian housing since the late 90s has been on an uninterrupted tear (yes, even in ‘08 and ‘20).

Huh?

The smartphone improvements hit a rubber wall a few years ago (disregarding folding screens, that compose a small market share, improvement rate slowed down drastically), and the industry is doing fine. It's not growing like it use to, but that just means people are keeping their smartphones for longer periods of time, not that people stopped using them.

Even if AI were to completely freeze right now, people will continue using it.

Why are people reacting like AI is going to get dropped?

People are dumping billions of dollars into it, mostly power, but it cannot turn profit.
So the companies who, for example, revived a nuclear power facility in order to feed their machine with ever diminishing returns of quality output are going to shut everything down at massive losses and countless hours of human work and lifespan thrown down the drain.
This will have an economic impact quite large as many newly created jobs go up in smoke and businesses who structured around the assumption of continued availability of high end AI need to reorganize or go out of business.
Search up the Dot Com Bubble.
People pay real money for smartphones.
- People pay real Money for AIaaS as well..
Because novelty is all it has. As soon as it stops improving in a way that makes people say "oh that's neat", it has to stand on the practical merits of its capabilities, which is, well, not much.
It’s absurdly unprofitable. OpenAI has billions of dollars in debt. It absolutely burns through energy and requires a lot of expensive hardware. People aren’t willing to pay enough to make it break even, let alone profit
- Eh, if the investment dollars start drying up, they'll likely start optimizing what they have to get more value for fewer resources. There is value in AI, I just don't think it's as high as they claim.
- Training new models is expensive. Running them can be fairly cheap. So no
People differentiate AI (the technology) from AI (the product being peddled by big corporations) without making clear that nuance (Or they mean just LLMs, or they aren't even aware the technology has a grassroots adoption outside of those big corporations). It will take time, and the bubble bursting might very well be a good thing for the technology into the future. If something is only know for it's capitalistic exploits it'll continue to be seen unfavorably even when it's proven it's value to those who care to look at it with an open mind. I read it mostly as those people rejoicing over those big corporations getting shafted for their greedy practices.
- the bubble bursting might very well be a good thing for the technology into the future
  I absolutely agree. It worked wonders for the Internet (dotcom boom in the 90s), and I imagine we'll see the same w/ AI sometime in the next 10 years or so. I do believe we're seeing a bubble here, and we're also seeing a significant shift in how we interact w/ technology, but it's neither as massive or as useless as proponents and opponents claim.
  I'm excited for the future, but not as excited for the transition period.

is this where we get to explain again why its not really ai?

Nope, just where you divest your stocks like any other tech run.
He is writing about LLM mainly, and that is absolutely AI, it's just not strong AI or general AI (AGI).
You can't invent your own meaning for existing established terms.
- LLMs are AI in the same way that the lane assist on my car is AI. Tech companies, however, very carefully and deliberately play up LLMs as being AGI or close to it. See for example toe convenient fear-mongering over the "risks" of AI, as though ChatGPT will become Skynet.
I have to do similar things when it comes to 'raytracing'. It meant one thing, and then a company comes along and calls something sorta similar the same thing, then everyone has these ideas of what it should be vs. what it actually is doing. Then later, a better version comes out that nearly matches the original term, but there's already a negative hype because it launched half baked and misnamed. Now they have to name the original thing something new new to market it because they destroyed the original name with a bad label and half baked product.

The hype should go the other way. Instead of bigger and bigger models that do more and more - have smaller models that are just as effective. Get them onto personal computers; get them onto phones; get them onto Arduino minis that cost $20 - and then have those models be as good as the big LLMs and Image gen programs.

Other than with language models, this has already happened: Take a look at apps such as Merlin Bird ID (identifies birds fairly well by sound and somewhat okay visually), WhoBird (identifies birds by sound, ) Seek (visually identifies plants, fungi, insects, and animals). All of them work offline. IMO these are much better uses of ML than spammer-friendly text generation.
- those are all classification problems, which is a fundamentally different kind of problem with less open-ended solutions, so it's not surprising that they are easier to train and deploy.
- Platnet and iNaturalist are pretty good for plant identification as well, I use them all the time to find out what's volunteering in my garden. Just looked them up and it turns out iNaturalist is by Seek.
This has already started to happen. The new llama3.2 model is only 3.7GB and it WAAAAY faster than anything else. It can thow a wall of text at you in just a couple of seconds. You're still not running it on $20 hardware, but you no longer need a 3090 to have something useful.
Well, you see, that's the really hard part of LLMs. Getting good results is a direct function of the size of the model. The bigger the model, the more effective it can be at its task. However, there's something called compute efficient frontier (technical but neatly explained video about it). Basically you can't make a model more effective at their computations beyond said linear boundary for any given size. The only way to make a model better, is to make it larger (what most mega corps have been doing) or radically change the algorithms and method underlying the model. But the latter has been proving to be extraordinarily hard. Mostly because to understand what is going on inside the model you need to think in rather abstract and esoteric mathematical principles that bend your mind backwards. You can compress an already trained model to run on smaller hardware. But to train them, you still need the humongously large datasets and power hungry processing. This is compounded by the fact that larger and larger models are ever more expensive while providing rapidly diminishing returns. Oh, and we are quickly running out of quality usable data, so shoveling more data after a certain point starts to actually provide worse results unless you dedicate thousands of hours of human labor producing, collecting and cleaning the new data. That's all even before you have to address data poisoning, where previously LLM generated data is fed back to train a model but it is very hard to prevent it from devolving into incoherence after a couple of generations.
- this is learning completely the wrong lesson. it has been well-known for a long time and very well demonstrated that smaller models trained on better-curated data can outperform larger ones trained using brute force "scaling". this idea that "bigger is better" needs to die, quickly, or else we're headed towards not only an AI winter but an even worse climate catastrophe as the energy requirements of AI inference on huge models obliterate progress on decarbonization overall.
That would be innovation, which I'm convinced no company can do anymore.
It feels like I learn that one of our modern innovations was already thought up and written down into a book in the 1950s, and just wasn't possible at that time due to some limitation in memory, precision, or some other metric. All we did was do 5 decades of marginal improvement to get to it, while not innovating much at all.
- Are you talking about something specific?

"The economics are likely to be grim," Marcus wrote on his Substack. "Sky high valuation of companies like OpenAI and Microsoft are largely based on the notion that LLMs will, with continued scaling, become artificial general intelligence."
"As I have always warned," he added, "that's just a fantasy."

Even Zuckerberg admits that trying to scale LLMs larger doesn’t work because the energy and compute requirements go up exponentially. There must exist a different architecture that is more efficient, since the meat computers in our skulls are hella efficient in comparison.
Once we figure that architecture out though, it’s very likely we will be able to surpass biological efficiency like we have in many industries.
- That's a bad analogy. We weren't able to surpass biological efficiency in industry sector because we figured out human anatomy and how to improve it. It's simply alternative ways to produce force like electricity and motors which had absolutely no relation to how muscles works.
  I imagine it would be the same for computers, simply another, better method to achieve something but it's so uncertain that it's barely worth discussing about.

Because nobody could have possibly saw that coming. /s

I think I've heard about enough of experts predicting the future lately.

Marcus is right, incremental improvements in AIs like ChatGPT will not lead to AGI and were never on that course to begin with. What LLMs do is fundamentally not "intelligence", they just imitate human response based on existing human-generated content. This can produce usable results, but not because the LLM has any understanding of the question. Since the current AI surge is based almost entirely on LLMs, the delusion that the industry will soon achieve AGI is doomed to fall apart - but not until a lot of smart speculators have gotten in and out and made a pile of money.

I am so tired of the ai hype and hate. Please give me my gen art interest back please just make it obscure again to program art I beg of you

It's still quite obscure to actually mess with AI art instead of just throwing prompts at it, resulting in slop of varying quality levels. And I don't mean controlnet, but github repos with comfyui plugins with little explanation but a link to a paper, or "this is absolutely mathematically unsound but fun to mess with". Messing with stuff other than conditioning or mere model selection.
- I know, it's actually still a beautiful community but much harder to talk to outsiders about

This is why you're seeing news articles from Sam Altman saying that AGI will blow past us without any societal impact. He's trying to lessen the blow of the bubble bursting for AI/ML.

It's been 5 minutes since the new thing did a new thing. Is it the end?

As I use copilot to write software, I have a hard time seeing how it'll get better than it already is. The fundamental problem of all machine learning is that the training data has to be good enough to solve the problem. So the problems I run into make sense, like:

Copilot can't read my mind and figure out what I'm trying to do.
I'm working on an uncommon problem where the typical solutions don't work
Copilot is unable to tell when it doesn't "know" the answer, because of course it's just simulating communication and doesn't really know anything.

2 and 3 could be alleviated, but probably not solved completely with more and better data or engineering changes - but obviously AI developers started by training the models on the most useful data and strategies that they think work best. 1 seems fundamentally unsolvable.

I think there could be some more advances in finding more and better use cases, but I'm a pessimist when it comes to any serious advances in the underlying technology.

Not copilot, but I run into a fourth problem:
4 The LLM gets hung up on insisting that a newer feature of the language I'm using is wrong and keeps focusing on "fixing" it, even though it has access to the newest correct specifications where the feature is explicitly defined and explained.
- Oh god yes, ran into this asking for a shell.nix file with a handful of tricky dependencies. It kept trying to do this insanely complicated temporary pull and build from git instead of just a 6 line file asking for the right packages.
- I've also run into this when trying to program in Rust. It just says that the newest features don't exist and keeps rolling back to an unsupported library.
So you use other people's open source code without crediting the authors or respecting their license conditions? Good for you, parasite.
- Very frequently, yes. As well as closed source code and intellectual property of all kinds. Anyone who tells you otherwise is a liar.
- Programmers don't have the luxury of using inferior toolsets.
- Ahh right, so when I use copilot to autocomplete the creation of more tests in exactly the same style of the tests I manually created with my own conscious thought, you're saying that it's really just copying what someone else wrote? If you really believe that, then you clearly don't understand how LLMs work.
- I completely understand where you’re coming from, and I absolutely agree with you, genAI is copyright infringement on a weapons-grade scale. With that said, though, in my opinion, I don’t know if calling people parasites like this will really convince people, or change anything. I don’t want to tone police you, if you want to tell people to get fucked, then go ahead, but I think being a bit more sympathetic to your fellow programmers and actually trying to help them see things from our perspective might actually change some minds. Just something to think about. I don’t have all the answers, feel free to ignore me. Much love!
Copilot can't read my mind and figure out what I'm trying to do.
Try writing comments

Welcome to the top of the sigmoid curve.

If you were wondering what 1999 felt like WRT to the internet, well, here we are. The Matrix was still fresh in everyone's mind and a lot of online tech innovation kinda plateaued, followed by some "market adjustments."

I think it's more likely a compound sigmoid (don't Google that). LLMs are composed of distinct technologies working together. As we've reached the inflection point of the scaling for one, we've pivoted implementations to get back on track. Notably, context windows are no longer an issue. But the most recent pivot came just this week, allowing for a huge jump in performance. There are more promising stepping stones coming into view. Is the exponential curve just a series of sigmoids stacked too close together? In any case, the article's correct - just adding more compute to the same exact implementation hasn't enabled scaling exponentially.

I work with people who work in this field. Everyone knows this, but there's also an increased effort in improvements all across the stack, not just the final LLM. I personally suspect the current generation of LLMs is at its peak, but with each breakthrough the technology will climb again.

Put differently, I still suspect LLMs will be at least twice as good in 10 years.

Of course it'll crash. Saying it's imminent though suggests someone needs to exercise their shorts.

I just want a portable self hosted LLM for specific tasks like programming or language learning.

You can install Ollama in a docker container and use that to install models to run locally. Some are really small and still pretty effective, like Llama 3.2 is only 3B and some are as little as 1B. It can be accessed through the terminal or you can use something like OpenWeb UI to have a more "ChatGPT" like interface.

I hope it all burns.

Seems to me the rationale is flawed. Even if it isn't strong or general AI, LLM based AI has found a lot of uses. I also don't recognize the claimed ignorance among people working with it, about the limitations of current AI models.

while you may be right, one would think that the problem lies in the overestimated peception of the abilities of llms leading to misplaced investor confidence -- which in turn leads to a bubble ready to burst.
- Yup. Investors have convinced themselves that this time AI development is going to grow exponentially. The breathless fantasies they’ve concocted for themselves require it. They’re going to be disappointed.
Can you name some of those uses that you see lasting in the long term or even the medium term? Because while it has been used for a lot of things it seems to be pretty bad at the overwhelming majority of them.
- AI is already VERY successful in some areas, when you take a photo, it is treated with AI features to improve the image, and when editing photos on your phone, the more sophisticated options are powered by AI. Almost all new cars have AI features.
  These are practical everyday uses, you don't even have to think about when using them.
  But it's completely irrelevant if I can see use cases that are sustainable or not. The fact is that major tech companies are investing billions in this.
  Of course all the biggest tech companies could all be wrong, but I bet they researched the issue more than me before investing.
  Show me by what logic you believe to know better.
  The claim that it needs to be strong AI to be useful is ridiculous.

so long, see you all in the next hype. Any guesses?

AI vagina Fleshlight beds. You just find your sleep inside one and it will do you all night long! Telling you stories of any topic. Massaging you in every possible way. Playing your favorite music. It's like a living room! Oh I'm sleeping in the living room again. Yeah I'm in the dog house. But that's why you need an AI vagina Fleshlight bed!
- Get a few more hours of sleep
Tradwives

Nice, looking forward to it! So much money and time wasted on pipe dreams and hype. We need to get back to some actually useful innovation.

Sigh I hope LLMs get dropped from the AI bandwagon because I do think they have some really cool use cases and love just running my little local models. Cut government spending like a madman, write the next great American novel, or eliminate actual jobs are not those use cases.

Short on the AI stocks before it crash!

The market can remain irrational longer than you can remain solvent.
A. Gary Shilling

Fingers crossed.

nvidia at least sells shovels, they already made some real profit unlike openai

Oh nice, another Gary Marcus "AI hitting a wall post."

Like his "Deep Learning Is Hitting a Wall" post on March 10th, 2022.

Indeed, not much has changed in the world of deep learning between spring 2022 and now.

No new model releases.

No leaps beyond what was expected.

\s

Gary Marcus is like a reverse Cassandra.

Consistently wrong, and yet regularly listened to, amplified, and believed.

The tech priests of Mars were right; death to abominable intelligence.

That's a Space Grudgin'

Yay

Theres no bracing for this, OpenAI CEO said the same thing like a year ago and people are still shovelling money at this dumpster fire today.

Crash? Doesn't it have to be moving at all to crash?

Nvidia shares ..

It's gonna crash like a self driving tesla. It's gonna fall apart like a cybertrukkk.

It's had all the signs of a bubble for the last few years.

Ya AI was never going to be it. But I wouldn’t understate its impact even in its current stage. I think it’ll be a tool that will be incredibly useful for just about every industry

There aren't many industries where results that are correct in the very common case everybody knows anyway, a bit wrong in the less common case and totally hallucinated in the actually original cases is useful. Especially if you can't distinguish between those automatically.

yep Knew ai should die some day.

supermicro's accountants have just resigned 🤭

Great!! ....I don't what chatGPT to go anywhere, I use it every day and Google has become assss.

Until Open AI announces a new 5t model or something and then the hype refreshes

Well classic computers will always limited and power hungry. Quantum computer is the key to AI achieving next level

The only people who say this know nothing about quantum or computers
- I love the or in this sentence
Quantum computers are only good at a very narrow subset of tasks. None of those tasks are related to Neural Networks, AGI, or the emulation of neurons.

Just put another number behind it. Luddites won't know the difference.

Luddites weren't against new technology, they were against the aristocrats using new technology as a tool or excuse to oppress and kill the labor class. The problem is not the new technology, the problem is that people were dying of hunger and being laid off in droves. Destroying the machinery, which almost always they were the operators of when working on said aristocrat's factories, was an act of protest, just like a riot, or a strike. It was a form of collective bargaining.

🤷‍♂️ I only use local generators at this point,so I don't care.

I believe this about as much as I believed the "We're about to experience the AI singularity" morons.

Over-promising and under-delivering are normal - that goes for both directions. I think it's undeniable that machine learning / AI has been improving at a steady pace. Sure, the current cycle will reach its peak, but there will be another and another. I don't believe AGI is a matter of if, more a matter of when. Whatever technology they're using now is not it, but it may play a part and even if it doesn't become AGI, it will be one more thing to tick off and explore new roads.

As someone said "this is the worst AI will ever be". It'll keep getting better. We just have to be prepared for the next cycle and not be caught off-guard. But knowing humans, we are terrible at long-term planning and preparing. The next cycle will catch the majority with their pants down and we will be scrambling to pull them up while whatever it is was developed changes our world yet again.