The AI bubble argument misunderstands both bubbles and AI(danielmiessler.com) |
The AI bubble argument misunderstands both bubbles and AI(danielmiessler.com) |
2. Each model from GPT3.5 to GPT 4 to 5 has been 10x more expensive to train. However, we have long ago entered the era of diminishing returns where 4 was a big improvement from 3.5 but 5 was a mediocre improvement (and to some, a regression) from the previous model. That suggests, unless a miracle occurs, 6 will be extremely expensive to train to imperceptible improvements.
3. The latest "features" from OpenAI:
- study mode https://openai.com/index/chatgpt-study-mode/
- the rumor for an office suite https://www.computerworld.com/article/4021949/openai-goes-fo...
show that they are running out of ideas what to do.
4. Anthropic is settling on huge cases: https://news.ycombinator.com/item?id=45142885 while struggling with usage vs monetization: https://news.ycombinator.com/item?id=44715471
5. Anthropic are very early in their journey to be already scraping the barrel, yet they recently 180 on their ToS and admit that they, in fact, will be training on your chats (suggesting that they can't get more/quality data elsewhere): https://www.anthropic.com/news/updates-to-our-consumer-terms.
6. OpenAI and Anthropic are both connected to CoreWeave, Nvidia, Microsoft, Google. What is funny is that CoreWeave is also connnected/dependent on Nvidia, and so is Microsoft and Google. Those companies' stocks are a major part of the US economy at the moment. https://www.reuters.com/business/dominant-ai-trade-confronts.... "Meanwhile, the combined market cap of the 10 biggest AI plays, including Nvidia, Broadcom and Microsoft -- stood at $18 trillion, BCA said in a note last week. That amounts to about 33% of S&P 500 stock market capitalization, up from around 15% in late 2022, according to BCA."
10 companies in the same sector, comprising 33% of the S&P capitalization is a 100% bubble, and if anything, one of the biggest ever to exist. And I, for one, can't wait for it to pop so I don't have to read hype booster articles like this on the front page of HackerNews.
> The .com bubble was the belief that if you took your mid-ass business to the internet, you would instantly become rich. That is what popped.
And currently there's a belief in the market that if you just add AI to it it's going to the moon. I don't see how there's a difference here.
- Virtual Reality: big hype in the early 90s (arcades, movies like Lawnmower Man) through to use cases today like surgical training, aviation training
- Mobile video calls: hyped in early 2000's with 3G and pre-iPhone devices. Actually took off with 4G and 5G plus iOS and Android phones
- 3D printing: back in 2013 we were expecting "a 3D printer in every home" ... today valuable in industrial prototyping
Looking back at 2025 we'll be saying "Remember when they said everyone would lose their jobs to AI..."
The idea that these tools won't at all improve from where they are now isn't a widely held position.
Neither VR, mobile video calls or 3D printing were expected to radically change the entire work economy, if not bring about actual human-like intelligence. None of those three technologies were in the hands of a handful of ultra-valuable companies, that in turn pretty much all depend on a single American manufacturer of hardware. None were threatening to destroy the Internet as we know it, or the concept of truth and credibility our modern world rely upon.
VR going nowhere was a wet fart, AI going nowhere is gonna, in my opinion and hope, crash the entire tech economy that's been injecting high doses of the hopium in the long period of post-COVID stagnation and inflation.
Even if... one would think that a capitalist economy would do great with more and capable workers. One would think that more stuff would get done. Right?
I think there is a good chance that it will, in fact, shift millions towards unemployment. I am pro technology, yet technology in the hands of profit seekers will only be used to seek profits.
It happened during the agricultural revolution and during the industrial revolution. Millions of people were made unemployed by more efficient technology. Millions had to flee the country sides to then be thrown out of factories a few decades later, leading to slums and mass poverty. So many that the government had to enact more and more welfare programs like public schools, and food programs.
Capitalism is the only economic system that cannot handle more workers. For-profit production is not compatible with mass employment.
Almost like capitalism shoots itself in the foot and then forgets about it.
I think reality differs. Most countries have for profit production and most have mass employment. Maybe 95% employed and 5% unemployed but it generally muddles along. The masses always seem to vote for it unless they have communism imposed at gunpoint with walls to prevent them escaping.
Oh yes, this must be why chatbot companies announced $25 billion of new funding last week.
https://companiesmarketcap.com/juniper-networks/stock-price-...
The 'what' that pops, probably, IMHO, is the investment and resources ecosystem dumped, poured, thrown, redirected, etc. to AI.
Just like the .com boom, the costs and money thrown at AI are outsized and will 'pop" as it matures. A 'pop" and leveling down to something perhaps not so outrageous is coming.
As the author states, .com popped but left the Internet still around and expanding. AI is exactly that kind of bubble.
“AI will change how business is done” is not a bubble because every innovation changes how business is done, it’s just a matter of degree. Various forms of AI have been changing businesses for a decade at least.
Because something could very easily pop soon, but this article is just laying the groundwork for an argument that "it wasn't AI popping, that was the belief you could run a AI adoption program and fire half your staff popping."
At some point, what's the difference? Obviously LLMs are here to stay in some form or another.
What the stock market does with that is an entirely different animal and honestly I'm not optimistic.
Anybody telling you otherwise is lying to you.
Especially when the entire argument seems to hinge on the author's dislike for the word Bubble. The bubble that pops isn't the actual thing, it's the money that was propping the thing up...
Looking at recent YC batches, YC seems to believe so.
He says bubbles are false beliefs about reality that everybody figures out is false. The claim that "[AI] will displace tens or hundreds of millions of knowledge workers in the next 3-10 years" is that false belief inflating the bubble. Big tech and investors are dumping money into AI thinking at least a sizeable portion of the white collar workforce is going to be replaced with $2000-$4000/month subscriptions to AI agents (I am so sick of the term "agent") acting almost autonomously. We're hitting the wall with what's feasible with LLM scaling laws even if you're dumping countries GDP worth of money into making bigger and better models. Is spending $500 billion to $1 trillion dollars to get something that performs marginally better than the best LLM today worth it?
To me this is very simple: certain companies have invested billions upon billions in AI data centres that need to see a very high return on investment in the years ahead to avoid a “pop”. I think nvidia is also being valued is if it’s a certainty that we will continue to build many such data centres every year for the foreseeable future.
These expectations are now being faced with the reality that AI isn’t making us that much more productive, and the improvements in LLMs are slowing down, so it’s not clear there will be a continued will to pay expensive subscription for these AIs, and certainly not clear that companies are willing to accept higher prices that may be need to get ROI.
There’s another way this could go down: I have a sneaking suspicion that we can get 90% of the productivity benefits of LLMs with models that can run locally. A combination of an algorithm breakthrough, better tuned models, better software tooling around them, and better hardware in the hands of end users, may get us to that 90%. That could end up making a lot of AI companies redundant.
AI will one day transform the world but I do not believe modern AI will make that happen, that will have to wait for futuristic AI. I guess that makes AI not a bubble but I can still label modern, i.e. current, AI a bubble, right?
EDIT: The fact that I'm downvoted for saying that tells you that not only we're in a massive bubble, but the author knows that he's personally invested in the bubble getting bigger.
So of course he'll say "it isn't a bubble".
From the article:
> But if they think the bubble is the idea that AI will cause foundational change in business, the economy, and human work, then—in my opinion—they're just wrong
That's just the author's opinion, not a fact. People claiming it's a bubble disagree with this part. And they might be right. As far as I understand, no massive productivity gains have been detected yet. Some expected gains weren't there. Some kinds of work might benefit some, but so far, nothing that justifies the enormous investments.
It's not the AI itself that's the bubble, but the "use more AI" attitude. The idea that the AI is a genius and better than people. Every company doing something with AI because they need to be seen doing something with AI. That's the bubble.
After it pops, there will still be AI, and some companies will be wildly successful with it, just like many .coms survived the .com crash. But investors will finally see that it's not going to meet the inflated investor expectations.
Consider XKCD. Will I lose interest in Randall Munroe because an AI thing will auto-generate any number of XKCD-style comics? Of course not.
The proliferation of AI has been compared to the proliferation of polyester.
The internet was always going to disrupt the way we think about the discovery and purchase of goods and services at the bare minimum. Some saw that the role of the personal computer was evolving from just computation to communication and so on.. this stuff was obvious early on. Irrespective of the equity bubble at that time, the potential was real. It was just too early.
There is nothing of this sort happening with LLMs. Where do they fit? Nobody has a clear answer and I don't see a clear one emerging any time soon. If someone has a very clear and direct answer, please feel free to reply :=)
I think what pops is the false belief.
.com -> going online will save you AI -> adding a chatbot will save you
To me, those are what qualify as bubbles. And the other stuff is just overheating, disruption, and other effects.
Then you're inventing your own term. Bubble is shorthand for the Economics term "speculative bubble" or "economic bubble" and has described the behavior of many markets over the centuries. What collapses is asset prices, not ideas. It's definitional...just ask your friendly neighborhood economist.
Edit: Here, I went and found the origin for you (via wikipedia). The South Seas bubble spawned the term:
"The metaphor indicated that the prices of the stock were inflated and fragile – expanded based on nothing but air, and vulnerable to a sudden burst, as in fact occurred."
So, then, I guess, nothing is a bubble for you. For me, a more reasonable definition is a large build-up of capital that, at some point, snaps or pops, then sits at a level much lower than the original.
If you take something like that approach of the "bubble" then you can have a more charitable discussion. The AI bubble discussion then is about if the current levels of investment and capital being dumped into AI will stay at the same levels or have a more dramatic dip to be more of the future run-rate, then you can see how people see the parallels of other "bubbles" and actually have more of a discussion I think.
My sense is most view this concept of the "bubble" as the assumption.
One just need to ask one question: "is there a big amount of companies in a specific sector that are valued far above their returns". If that answer is yes, that means we are in a speculative bubble. The current insane valuations of unprofitable companies point to the largest speculative bubble ever to exist. I can't wait for it to pop faster already so there are less articles like this.
In many situations, the work is not indivisible. If AI can handle 80% of the work, then a company can let AI handle that 80%, fire 80% of their people, and consolidate the remaining 20% still-human-work with whoever is left.
That all being said I strongly disagree that GPT-5 isn't categorically better, I just think it's less obvious because we're starting to hit human cognitive ability to even assess that limits.
"is there a big amount of companies in a specific sector that are valued far above their returns". If that answer is yes, that means we are in a speculative bubble."
Pretty much.
What you're saying would equally apply to a kid annoyed at his dad because his friends have a Ferrari but his family only has a Porsche. It's childish.
Most of the world, like 90%+, is verifiably poor and struggling.
Every year the world economic organizations (all of em) pump out metrics about people coming out of poverty, but these are political and subject to bias.
Above 10 dollars a week is not poverty? Are you sure? 521 dollars a year?
And of course not all of the world is poor in the same capacity. Where a person can only eat meat once a week is different to where a person cant pay their rent, but they are both equally damaging and poverty.
What you think has been happening is not what has been happening.
In the Soviet Union and China most people were serfs or indentured servants. That is, the majority of the population.
They were bound by debt to serve either the state or local landlords.
Their revolutions werent acts of jealousy like you want to believe, they were real, spontaneous movements that came from the people. The communists only directed the movement that was already there.
Do not be childish, please read history seriously.
Capitalism has a flaw, it only does what is profitable.
What is profitable is not one and the same with what is socially necessary or even what is good. They're more like analogous.
Like I said before capitalism is the only economic system that cannot handle more workers. If you produce for profit you cannot produce for needs. In a sense profits vs needs are made of the same stuff but dont look the same. Especially at scale.
Employing as many people as possible wouldnt be a problem if you were working for needs because, simply, more needs would be covered.
Its literally that simple. And when there is a true surplus of workers it would mean less work/more leisure.
There are billions of us. We cannot all be capitalists and start our own businesses. Literally 100 years ago American socialists like Olive Johnson were already pointing out how the profit motive from large players has ruined mom and pop shops and made medium businesses almost completely beholden to production and finance monopolies.
What do you think will happen with the increased production from AI? Will the capitalist just allow the masses to compete openly with them?
I doubt it. Most likely more monopolization will occur.
Small business owners are the tip (retail) of the massive whole-sale industrial-production monopolies or they are artisans at best. And the masses are the rest of us, the 95%.
Western suburban mentality always posits that these ills cannot happen, even though they were continuously happening even through out the 20th century (Detroit is just one example).
And the "regulation" argument is very popular but I feel it ignores the real problem for us: there is no democracy.
With the regulation argument you're basically hoping that one of the two parties full of billionaires, that we explicitely do not control, shoot themselves in the foot.
And as to adding value to the chain, that is what workers currently do, thats why they get paid. Which is what sparked this argument. The economy is not infinitely flexible, not all will be able to adapt, and according to the rules of capital the adaptation will be competitive and exclusive, so many people will be left out.
Comments such as "Ai is inevitable" okay... post something tangible dude. All this stuff is just total noise.
Comment: AI isn't so bad
You: But LLMs are no good
I see that again and again with AI arguments. It's not a good argument really. Like saying aviation will fail because biplanes weren't very good.