- cross-posted to:
- usa@lemmy.ml
- cross-posted to:
- usa@lemmy.ml
You must log in or # to comment.
AI is funded solely by sunk cost fallacy at this point. I wonder how long it will be before investments start getting pulled back because of a lack of ROI. I can already feel the sentiment towards AI and it getting pushed in everything turning negative amongst consumers recently.
Investment is done really to train models for ever more miniscule gains. I feel like the current choices are enough to satisfy who is interested in such services, and what really is lacking is now more hardware dedicated to single user sessions to improve quality of output with the current models.
But I really want to see more development on offline services, as right now it is really done only by hobbyists and only occasionally large companies with a little dripfeed (Facebook Llama, original Deepseek model [latter being pretty much useless as no one has the hardware to run it]).
I remember seeing the Samsung Galaxy Fold 7 (“the first AI phone”, unironic cit.) presentation and listening to them talking about all the AI features instead of the real phone capabilities. “All of this is offline, right? A powerful smartphone… makes sense to have local models for tasks.” but it later became abundantly clear it was just repackaged always-online Gemini for the entire presentation on $2000 of hardware.
what really is lacking is now more hardware dedicated to single user sessions to improve quality of output with the current models
That is the exact opposite of my opinion. They’re throwing tons of computing at the current models. It has produced little improvement. The vast majority of investment is in compute hardware, rather than R&D. They need more R&D to improve the underlying models. More hardware isn’t going to get the significant gains we need
The problem is there is little continuous cash flow for on prem personal services. Look at Samsung’s home automation, its nearly all online features and when the internet is out you are SOL.
To have your own Github Copilot in a device the size and power usage of a Raspberry Pi would be amazing. But then they won’t get subscriptions.
more development on offline services
There is absolutely massive development on open weight models that can be used offline/privately. Minimax M2, most recent one, has comparable benchmark scores to the private US megatech models at 1/12th the cost, and at higher token throughput. Qwen, GLM, deepseek have comparable models to M2, and have smaller models more easily used on very modest hardware.
Closed megatech datacenter AI strategy is partnership with US government/military for oppressive control of humanity. Spending 12x more per token while empowering big tech/US empire to steal from and oppress you is not worth a small fraction in benchmark/quality improvement.
I wouldn’t have a problem if they were actually investing the money in something useful like R&D
Nearly all the investment is in data centers. Their approach for the past 2 years seems to be just throwing more hardware at existing approaches, which is a really great way to burn an absurd amount of money for little to nothing in return
It’s very corporate, isn’t it? “Just keep scaling what we have.”
That being said, a lot of innovation is happening, but goes unused. It’s incredible how my promising papers come out, and get completely passed over by Big Tech AI, like nothing matters unless it’s developed in house.
The Chinese firms are picking up some research in bigger models, at least, but are kinda falling into local maxima too.
One of our biggest bookstores contracted with a local artist for some merch. That artist used AI with predictable results. Now everyone involved is getting raked over the coals for it.
No surprise, they just announced a 4th round of layoffs too. 😟
https://www.koin.com/news/portland/powells-layoffs-employees-10292025/
Why do you think AI is pushed so hard?
Everyone is aware this has to be useful. Too much money.
Still the powers that be will do everything to avoid a hard crash, which would be so much earned.
When I lose $11 Billion dollars, I have to go to bed without supper.
If you owe the bank $100, that’s your problem; if you owe the bank $100 million, that’s the bank’s problem.
The difference between 100 million and 11.5 billion is about 11 billion. If you own a bank 11 billion that’s not only that bank’s problem, it’s the economy’s problem.
It’s about 11.5 billion, really.
It’s exactly 11.4 billion, really.
Well, if you want to get exact, sure. But if we’re talking about half units, like 11 and a half billion, then 11.4 is so close to 11.5 there’s no difference and calling it just about 11 sorta implies that it’s a more significant difference IMO
You need to be as precise as your resolution, otherwise the precision is meaningless. I guess you could argue that your resolution is units of half-billion (since some things are measured like that), but the initial value of 0.1B, and your use of 0.5 rather than ‘half’ suggests a resolution of 0.1B.
This is different to the aphorism ‘The difference between a million and a billion is about a billion’, both because of the difference in scale, and the quoted resolution.
I was referring to the general concept behind the quote.
I originally want to post the OG (apocryphal?) variant:
Owe Your Banker £1,000 and You Are at His Mercy; Owe Him £1 Million and the Position Is Reversed
But it sounds rather quaint these days.
is this $11,500,000,000 in real money or speculative money?
There is no difference anymore
What’s the difference?
You think it’s 11.5B worth of gold bars?
Maybe, just maybe, the bubble started bursting now.
Would be nice, I want to buy some ETFs at a discount
I look forward to the AI bubble bursting, and billionaires looking shocked, ‘because there were no signs’
They won’t lose any money…
In contrast to the housing bubble, where a lot of the value was in overpriced houses sold to individuals, this overpricing is almost entirely in tech stocks, and tech stocks are almost entirely owned by by the wealthiest 10%, even 1%. The tech billionaires have limited ability to divest themselves of their own overpriced companies and absolutely will lose money.
None of them are going bankrupt, they’ll all be just fine when the market recovers in a few years, because that’s the nature of capitalism. A bunch of peons, who convinced themselves that the bubble-value of their 401k meant it was safe to retire, will suffer, will have to go back to work - if you’re not an oligarch, losing money is painful.
In contrast to the housing bubble, where a lot of the value was in overpriced houses sold to individuals,
was?
34% of single-family housing are now rented out. We’re not building enough housing, a fact that hasn’t changed since the housing bubble collapsed. So, a lack of investment in new property, large funds putting money into existing properties instead, and less risky homeowners overall means we don’t really have a housing bubble. We have a supply shortage, leading to high prices. The correction for that isn’t a crash.
since the housing bubble collapsed
did it? it does not seem so. where I live to buy a house in good condition people need to take out loans that the bank may not even allow, but if it does they’ll pay it for decades. even empty plots are still very expensive. more and more people live in a rented place even though they don’t want to move, because their house is taken away.
Yeah all the people praying for a crash are praying for nobody to have retirement funds.
You can easily tell who’s actually employed in this thread because anyone with a 401k is going to get dicked down while the 0.1% get a bailout.
They will be fine either way…
A lot of them are actively talking about how it could be a bubble and the implications. No one is going to be surprised. Billionaires are just really hoping they can make it work before the bubble pops
Yeah… I mean billionaires have at least as much information as a random person on Lemmy :D
So they “lost” $11.5B? Cool, I lost 20 bucks last week and still had to explain it to my accountant 🤭 Feels like the entire AI industry is built on “don’t worry, growth will save us”, but at some point someone has to pay the electricity bill…
No, you don’t understand. Lose money on every sale, but make up for it in volume!
If you lose enough money, you trigger an integer overflow and end up with a profit!
Wow. Glad they just converted to a for profit entity! Can’t wait for them to unleash all this success on to the the general financial market.
This reminds me of something that came up recently. Copilot started hallicinating quite a bit more than usual in Copilot reviews. That made me think about the cost of operarion. As they burn money like this, I won’t be surprised if they start decreasing inference quality to decrease cost per user. Which also means people relying on certain model behaviour for tasks could get nasty surprises. Especially within automation workflows where model outputs aren’t being reviewed.
Copilot and Gemini are trash. They are driving away future business.
Github Copilot is somewhat useful for programming (or it feels useful when it cranks out some boring and routine code to my specs - not sure it actually saves me time though because I always review it all), but of course Microsoft have given a range of products all the same name for maximum confusion, as they do. The Copilot in Windows may be rubbish for all I know. I haven’t ever felt the need to press the button.
Anyone using something with inconsistent output in their automation deserves what they get.
Its a bubble
More like a drain amirite???
The whole “AI” thing is one big grift.
I agree, and essentially they used slightly reworked old neural network technologies, increasing their power with the help of data centers.
Oh no!
Anyway…
Microsoft also lowered stake in OpenAI from 32.5% to 27%
https://blogs.microsoft.com/blog/2025/10/28/the-next-chapter-of-the-microsoft-openai-partnership/the smaller the share, the smaller the quarterly losses. Other sources have included that MSFT now gets a 20% royalty on openAI revenue, but its not in that PR. It’s not clear why else MSFT share would have fallen.
apparently the bubble might not be as extreme as some people think because the major AI players are all being propped up by companies that actually produce revenues and profits
You know if you invest all your winnings into all the companies that buy your stuff so that they can buy more of your stuff, you are actually not generating any winnings.
Like Nividia which… Oh all based on AI revenue.
And even though NVIDIA is better place as they do produce something, but the something in play has little value out of the AI bubble.
NVIDIA could be left holding the bag on a super increased capacity to produce something that nobody wants anymore (or at least nowhere near at the levels we have now) so they are still very much exposed.
I want cheap GPUs at home please!
me too, but the GPU used for AI are not the same as what we would use at home.
maybe the factories can produce both kinds and they would be cheaper, but it is speculation at this point
It’s literally the same chip designers, production facilities and software. Every product using <5nm silicon fabs compete for the same manufacturing capabilities (fab time at TSMC in Taiwan) and all Nvidia GPUs share lots of commonalities in their software stack.
The silicon fab producing the latest Blackwell AI chips is the same fab producing the latest consumer silicon for both AMD, Apple, Intel and Nvidia. (Let’s ignore the fabs making memory for now.) Internally at Nvidia, I assume they have shuffled lots and lots of internal resources over from the consumer oriented parts of the company to the B2B oriented parts, severely reducing consumer focus.
And then we have any intentional price inflation and market segmentation. Cheap consumer GPUs that are a bit too efficient at LLM inference will compete with Nvidias DC offerings. The amount of consumer grade silicon used for AI inference is already staggering, and Nvidia is actively holding back that market segment.
They’re the ones selling shovels in this gold rush, though.
The problem is that the companies that actually produce revenues and profits are also in turn being propped up by AI.
such as?
Who wants to give me a billion dollars to dig a hole and I’ll give you a billion to fill it back in and we’ll both say to investors we posted a billion dollars in revenue.
Finally, some good news.
They will just wait it out until teens will not be able to use a search engine, because they grow using free GPT and as soon as those earn enough money, they’ll start the enshitification.
Of course, teachers and students will have still free access, do that they not learn to find stuff not with AI.
they will somehow shift the cost on the consumer and continue on
