For PC gaming news and discussion. PCGamingWiki
Rules:
The thing about market cap is that it's kinda not real. You think Rivian was ever worth 130 billion dollars, for example? Nah.
Nvidia has only lost about the last three months of stock gains. It's quite bad, and they'll cry, but they'll be fine in the long term.
"Long term? But I want all the money now!"
This is such a concise and accurate description of everything that's wrong with corporations.
Or at least publicly traded corporations
They really need to change how market cap is calculated. Even a linear function going to zero as the last share is sold would be better. I wonder.
I think it would make more sense to have good journalism and honest coverage of markets, and an informed, financially literate populace. While I'm at it I want a pony who cleans the dishes and shits sunglasses.
Well yeah. The market cap is all the stock x the current market price. But if even a few % of people sell then the price will drop. You can never sell everything at the top price.
It's up 5% in premarket right now, so... We will see how long they stay down.
Pretty typical that the only ones to suffer any lasting harm will be the longs and shorts who weren't first to act.
Yeah. I think long term, many advancements in Ai will still need Nvidia chips and it will be better to have more of them still.
I mean, stock price rarely correlates with actual need and more with just hype. If you expect many people to buy a given stock, it makes sense for you to buy before them. If you expect many people to sell a given stock, it makes sense for you to sell before them. The actual need kind of just provides a baseline, i.e. even if the hype dies off completely, it'll still make some profit and pay out some fraction from that to anyone who's willing to park their money there.
As to genuine AI needing better GPUs, I'll believe it when Jeff Hawkins says it's so.
Why is that?
I've read a couple of his books, and he's been developing a theory of human intelligence with the express goal of creating true AGI. His Thousand Brains book espouses a theory I find just about credible and, perhaps eventually, actionable.
It's all to do with cortical columns working in parallel on a pseudo-voting type of system. If at some point he says "we know how to make AGI now, but we're going to need really good GPUs", I will find that credible based on having read his work.
Yeah. I dont think what we have now is general intelligence. It's just word prediction models. But they still manage to give an impression of intelligence. I think there will be some big change at some time where Ai models will use some different technology and probably give something closer to actual thinking.
Oh, nobody thinks what we have now is AGI. Not even the AI thinks it's AGI. It's just a words firehose being hyped to hell and gone because billionaires are so excited to fire their employees that they're jumping the gun.
Or said billionaires are trying to squeeze more money out of the AI bubble
Oh noes! 600 billion fluffybucks that only ever existed in the feverish fantasies of stock market analysts! Oh the humanity!
Meh, this is how Tech has always worked. The list of companies that foresaw endless growth due to a temporary market advantage is long. As an example ask Intel how their DRAM production is looking these days. They were dominating the market with it in early 70s but by the early 80s they'd been entirely supplanted. Anybody still buying new ZIP drives? How much market relevance does MySpace or Napster have these days?
This particular stock "crash" is really just an investor driven blip that Nvidia will quickly put in its rearview mirror but its a signal flare that the AI Market in general and the GPU market in particular is ripe and waiting for disruption. That disruption is coming and it will arrive quicker than most realize.
They are already roughly half way back.
It's okay, it's only pretend money.