10 years? Ain’t nobody got time for that

Well duh. That’s how training an LLM works

I assume there’s a bottom at some point. Bedrock or something like that. But idk I’m not a geologist.

I guess the “lucky” part is that it’s sinking evenly. Otherwise buildings would be collapsing all the time.

Well, if he wants them, and is going to pay for transport… Definitely better than killing them. I don’t think Colombians would mind if he took all 160.

It’s an open weights model - you can run your own. If you’re lucky enough to have a GPU or two laying around … then it can be it’s a lot cheaper actually.

“starting?” I assumed everyone who wasn’t onboard left years ago.
At least Linkedin is full of “ex-Palantir” people, but tbh Linkedin job descriptions are not know for their truthfulness.

Firing 30,000 of them. You can’t fire people if you don’t hire them in the first place. And as you all know stock price jumps after big layoffs.

deleted by creator

I think it’s about video generation. They shut down Sora because it was burning too much money. Imagine how much a massive amount of porn videos would cost.

No your honor, I didn’t shoot that guy, the gun did it

Also a fairly high suicide rate. Self-selection for happiness?

Venezuelans will get fuck all from the deal, but they also got nothing before either. From zero to zero. At least they are not losing.

The biggest surprise in the article is that Atlassian is not profitable. How? They pretty much have a monopoly in the Jira-like space (look , I can’t even think of a generic name) and they charge a hefty sum for their products. How tf do they lose money?

I’m not surprised. Jira is a monster.
As with any software, look how complex it is as a user, the hidden part is 10x worse.

LLMs are predictive text machines. Focus on “predictive”. Of course they will not output random text.

Note: not fully deterministic though - they need (pseudo)randomness at few critical points to be good

Look at it from the bright side. Manufacturers are building massive new capacity for demand that will never come. Already produced chips can’t be repurposed but machinery can, easily. In a few years RAM will be dirt cheap.

They used to sell a pretty good (if complex) database system. However it hasn’t been popular for many years. I assume they still have big customers who are locked in.

These days they’re just another amorphous “cloud service provider”, and not a good one either.

How do you even achieve that? I have to coax it into correctly running the project locally.

To the surprise of absolutely nobody