Originally published at: https://boingboing.net/2024/06/19/nvidia-is-the-worlds-largest-publicly-traded-company.html
…
7 Likes
9 Likes
huge wobbling soap bubble that floats menacingly over the audience for a few seconds before popping wetly over their heads. But perhaps I’m just …. swipes shades …. feeling invidious.
Is our greatly esteemed poster-in-general suggesting that “A.I.” and cryptocurrency shall go the way of NFT bored-apes and tulip bulbs!? tune in next week when we’ll learn exactly when we should’ve shorted. (snerk upon the capt’n “invidious”)
8 Likes
At least tulips don’t put a huge drain on power and water supplies.
10 Likes
At least until this AI bubble bursts…
7 Likes
That company’s growth is incredible, and in the rearview mirror, obvious.
I was doing Xray tomography three years or so ago. This wasn’t medical or even more interesting standard industrial stuff - we used mostly homebrew software to do crazy setups and very high energy imaging (15 MeV).
We begged for every one of their best processor cards to munch the ~TB datasets but they were unobtanium thanks to the crypto thing. Even with any pull we could swing we couldn’t get more than a trickle of the goods.
2020-me really should have thought harder on the sexy (then?) GPU ecosystem and invested deeply…
8 Likes
… meanwhile in the real world
4 Likes
3 Likes
Hypothetically the bubble bursts and the stock is worth a tiny fraction of what it is today. Does that mean boingboing stops using AI generated images? Will we all stop playing games with 3D graphics?
Or simply NVIDIA continues doing what it is doing, making increasing revenue. And the stock’s P/E falls to some more reasonable level. Tough for investors but kind of irrelevant for NV’s business.
1 Like
Not like this is a BB specific problem… And just because a “technology” rolls out, doesn’t mean it’ll stick around, especially when it’s generally speaking a proven failure in most of the things it does… When was the last time you watched a laserdisc or Beta max or an 8-track…
Yes, profits ALWAYS go up… there has never been a time when stuff crashes… Especially not in 1929… /s
I can’t predict the future. I can only comment on data of previous quarters.
But not everyone did poorly during the Great Depression. William Boeing and Walter Chrysler did exceptionally well.
The take away, I guess, is if the market crashes, be the one flexible enough to bounce between crypto, machine learning, compute, gaming, etc. Rather than an AI startup that has their entire business invested in developing a stack for a market that may disappear one day.
It’s not like the market for videos and portable music went away. The technology gets replaced with an equivalent as long as there is a need and profit to be had for that need. And oh my has there been a lot of profit selling movies and music to play at home.
Are LLMs going to be around in 20 years? Probably not. It will be something new that fills the same need for generative machine learning.
And I think most chip vendors realize this, because I’m not aware of any of them that are hardwired to only do LLMs. The ones I’m familiar with handle some specialization for “AI” by handling some operations for deep neural networks (DNNs) and Convolutional neural networks (CNNs). Usually based around how to handle memory virtualization and non-linear walks needed for operations over a large matrix. Rather than anything overly specific to a particular algorithm. Other features for training versus inference have to do with the support and performance of higher precision numbers. You want to train with a high precision, but inference doesn’t need the resolution and is faster and lower energy when they use fewer bits.
Oh, that’s okay then… I mean, after all, what matters are the quarterly earnings, not the emiseration of millions…
I don’t care about any of that. I care about having a peoeple-centered economy that focuses on what we actually need to thrive as people, not what will make shithead CEOs wealthy beyond belief.
Principle is the same. Technology is never a straight line. And we might do better to focus on what will humantiy, rather than what will make a few wealthy.
I think this pretty much sums up what I think of AI at this point…
https://ludic.mataroa.blog/blog/i-will-fucking-piledrive-you-if-you-mention-ai-again/
2 Likes
Agreed. Nothing in my previous comments suggested otherwise.
But what you’re saying now is different from the earlier suggesting that AI is a bubble. I suppose both could be true, but I feel like the subject has shifted.
Ultimately I want to know two things:
- What happens if/when the AI bubble “pops”
- Is this a bubble and inevitable, or is this technology going to be pervasive in our daily lives like cellphones?
Both of these question I ask not even from a money-making standpoint but how does it affect average people who don’t have millions. The cynic in me suspects: Nothing good comes from this for anyone below a certain income level.
The support of the concentration of wealth via the stock market is directly at odds with the welfare of humanity. AI (and all the many tech bubbles since the 90s) are generally speaking directly at odds with the welfare of humanity. These bubbles have super-charged inequality and we’re seeing how it’s impacting climate change.
We can certainly have healthy innovation in the tech sector that benefits all of us without all the bullshit hypecycle that is warping the economy by the day, in favor of the already wealthy.
The same thing that happens since the 80s with these bubbles (tech or otherwise) the losses are socialized, and we’re all on the hook for them.
Yes. Everyone I trust to discuss tech and the economy agrees it is a bubble and it will burst (people who aren’t easily swayed by bullshit like Doctorow or folks like Bruce Schneier.
But it IS, because it’s about the stockmarket, not about remotely about how technology impacts people’s daily lives outside of money…
See above.
These are bullshit hypecycles where, as noted in the article I posted, CEOs get invested, push this bullshit against the advice of their technology teams, until it crashes and the new shiny comes along, and the cycle repeats itself. I know this, because I’ve been hearing this same cycle of new technologies from my better half who has been programming professionally for decades now. This is how the modern, neo-liberal economy, freed from its new deal keynsian restraints work. 
To some degree, this has been doing on since the dawn of the capitalist age (see Brian Merchant’s book on the Luddites, for example). So it’ snot just me whinging about AI, it’s a shown historical pattern inherent to the capitalist system.
2 Likes
It is for me.
This topic was automatically closed after 5 days. New replies are no longer allowed.
