Many companies in the US (and in Europe to some extent) that were the AI speculation/hype winners got a lot of value and capital on the notion that AI progress meant spending hundreds of billions every year to expand data centers and buy way more chips. Deepseek outperformed proprietary models in certain metrics and performed comparably otherwise, and was built using open source tools, is open source itself, and cost several orders of magnitude less. This, on top of none of it being profitable, is likely to drive investors to look elsewhere because they’ll rightfully ask “what the fuck are we paying you for?”
5 Likes
the music is about to stop.
2 Likes
I love Dan Olsen
3 Likes
one of hte best video essayist. mantracks fucking rules as a video
2 Likes
Holy carp
is jon bois video essay
discuss
2 Likes
the line between video essayist and documentarian is nonexistent
2 Likes
Yeah. And they did it using lower end chips which is why Nvidia cratered. On Monday there was about $1.2 trillion lost in market cap across all the tech companies, the very same ones I was saying the other day were likely going to take a hit soon which is a nice little feather in my cap I suppose (I just listened to people that were smarter than me).
4 Likes
disagree
there’s a line but it’s insanely fuzzy
Huh that explains why everyone’s freaking out about it.
1 Like
anyways we should organize a twilight imperium night
thoughts?
s&p500 is tech-heavy and is still positive over the last week so as of right now selling everything still a bad move unless your life savings were in NVIDIA lmao
1 Like
I would agree, if most of the site wouldn’t liive a different timezone thn me…
im sure at some point the market will dip
but then it’ll go back up (unless we have bigger problems LOL)
trying to predict and time the market is just way riskier than holding a well-diversified portfolio in the long run 
1 Like
It’s down a full point in last 5 days after a bit of an upward bounce?
1 Like
I’m still wondering if we’re going to get robots that can think like humans in their own ways of finding about the world without depending on a literature base. Oh and if emotions will ever be possible
(This goes for any board game above 1 hour in average playtime.)
my retirement savings accounts are all up and they’re literally 100% in s&p500 funds so that seems weird
1 Like
1 Like
i suspect we’re looking at slighly different ranges with 1wk vs 5d
1 Like