Michael Burry of "The Big Short" fame warns of inflated tech stock valuations amid fears of an AI bubble

[takemaru]

False. That means it would take 40 years if NVidia's earnings remained constant. Yet their projected earnings are expected to rise drastically over the next few years.

Their earnings report will come out if a few days. I wouldn't bet on their earnings going up, most companies/corporations have already loaded up on GPUs or developed an ASIC for the job.

 

[LimyG]

False. That means it would take 40 years if NVidia's earnings remained constant. Yet their projected earnings are expected to rise drastically over the next few years.

Thanks Endymio. No need for me to add anything, just quoted your post.

 

[LimyG]

Their earnings report will come out if a few days. I wouldn't bet on their earnings going up, most companies/corporations have already loaded up on GPUs or developed an ASIC for the job.

Personally, I don't know. I'm not a stocks and shares specialist.
But, over all. A.I. isn't going away. Heck anthropic made a new version of Claude (sorry forgot the name), that's so powerful they didn't realease it to the general public. Not yet anyway.

I'm no fan of A.I because of Microsofts forcing of it. But that is just sentiment, the rest of this post is not about how I feel about it, it's the way I see it going forward:

A.I has already for instance drastically improved early detection of Breast Cancer. Things that even the best doctors would miss. That's just one example which has already changed peoples lifes, literally. (Mostly women I would assume, ahem.)

There are many more examples and A.I is still very much in it's infancy. People are still learning how to use it in ways that will affect, probably, everything. Directly or indirectly. I don't pretent to know what exactly, and with great power there are always risks (bad actors using it for bad things) but that's life, and won't hold it up. Legislation won't keep up either.

I did used to think there was an A.I. bubble, but while adjustments are likely, the power of it
(to do both good and bad - in the wrong hands) is a defining moment in the 21st century.
Sounds dramatic? That's because it is!

TBH the power of A.I (not talking about windows) has surprised, even shocked me recently, after reading about it's applications in science/medicine. Again I want to stress we are in the very infancy of A.I. as of 2026. The cats out of the bag, and it's only just starting.

Sure there is talk of circular investments, over investments and other negative indicatiors, this could cause something "lesser," to burst. But we are not talking about one company here. Even in NV were to go bankrupt, it's not going stop A.I.

The sheer potential, all encompassing possiblities of A.I. the fact it's not a single thing or fad, and that it can be applied and used in, well more things than I can imagine - it's HUGE.

It's for those reasons, the sheer (and to me) unknown potential of it, that it is simply not something that will "burst."

Finally my opinion, or if I like or hate A.I is irrelevant. It's arrived, it's accelerating, and no one can stop it.

 

[takemaru]

Thanks Endymio. No need for me to add anything, just quoted your post.

Their earnings are expected to rise.

That assumes investors will just keep pumping money into the sector at an increasing rate quarter by quarter. Which is not impossible, but at some point investors will either run out of money or patience. AI models are not making any money right now, they are subsidised by corporations in an attempt to corner the market.

https://www.tomshardware.com/tech-i...hips-sitting-in-inventory-that-I-cant-plug-in

Chinese startup founded by Google engineer claims to have developed its own TPU chip for AI — custom ASIC reportedly 1.5 times faster than Nvidia's A100 GPU from 2020, 42% more efficient

It's allegedly 1.5 times the speed of Nvidia's A100 GPU from 2020.

www.tomshardware.com

Nvidia sales are 'off the charts,' but Google, Amazon and others now make their own custom AI chips

Nvidia is king in AI chips, but custom ASICs are gaining ground with Google, Amazon, Meta, Microsoft and OpenAI now making their own.

www.cnbc.com

 

[Endymio]

Their earnings are expected to rise.

That assumes investors will just keep pumping money into the sector at an increasing rate quarter by quarter. Which is not impossible, but at some point investors will either run out of money or patience. AI models are not making any money right now, they are subsidised by corporations in an attempt to corner the market.

Too many false assumptions here. First, NVidia's profits can easily rise even against static revenues, as they amortize R&D through the product line. Second and far more relevant is that total global spending on AI is still in its infancy, and isn't simply expected "to rise somewhat", but increase by one or even two orders of magnitude over the next couple of decades. Third and worst of all is the absurd belief that "all AI models aren't making any money". I am personally invested in three different AI firms, all of which are not only making money, but showing gross margins above 50%.

Your one relevant point is that NVidia is facing increasing competition. It's difficult to predict what that will mean long term ... but in the short term, NVidia has done an amazing job at holding those competitors off, and continuing to show increasing growth.

 

[bellgrand]

Utter nonsense. First of all, literally tens of thousands of companies are profiting off AI, even though they're not "AI firms". And even AI-only corporations are doing very well -- I'm personally invested in three which are turning a large profit, two of the three have gross margins above 50% ....an absolutely insane level.

Total AI investment in 2026 is expected to be in excess of $650 billion. That's billion with a "B".

Most recent business surveys report no significant value being produced by AI spending.

Gross margins are high, sure. But gross margins do not include capex. So while OpenAI has a gross margin of 33%, their operating margin once you account for spend is projected to be -69% this year.

 

[Endymio]

Most recent business surveys report no significant value being produced by AI spending.

You've been misled by bad journalism. First of all, these reports are based on surveys. Statistically, an opt-in survey is useless, and opt-in suveys where participants have a strong incentive to lie are worse than useless.

If I was the CEO of a business seeing large gains from adoption of AI, the first thing I'd do when receiving one of those survey links is log on and tell all my competitors it's been a disaster, and they should avoid it at all costs.

The real indicator here is adoption and spending. Companies adopting AI tools continue to do so, and increase their spending on same. And those companies are now showing record profits.

Gross margins are high, sure. But gross margins do not include capex.

Err, most businesses consuming AI products are reporting the expense as OpEx, not CapEx ... and it isn't just gross margins that are high, but net margins and -- most importantly -- profits.

So while OpenAI has a gross margin of 33%, their operating margin once you account for spend is projected to be -69% this year.

Was this a joke? OpenAI is attempting to become another Amazon, which showed negative margins for two full decades --- in order to finance the growth that turned it into a three trillion dollar profitable firm.

Not that I think OpenAI will see any similar success. But it's merely one company among hundreds providing AI products -- and the real touchstone is the success of those 100,000+ companies consuming the products.