Hacker Newsnew | past | comments | ask | show | jobs | submit | more htrp's commentslogin

>Price changes are coming exclusively to Dell's commercial business, selling machines to companies and corporate clients [...] starting dec 17

>His thesis was sound in a rational world but our world isn't rational anymore and his data center REIT short got obliterated by the AI bubble.

I suspect he might be a bit bitter about that one. He has also shorted tesla (rightly or wrongly) as well as Carvana


His Tesla short was off many years before his fund closed and he pretty much said he wasn't going to bet in meme stocks since their value is so detached from any kind of objective analysis.

The problem for him (and muddy waters and other shorts) was the memification of the entire economy. If you listen to his data center REIT short thesis it was very sound. Data center REITs were not meme stocks and there was every reason for them to respond to economic gravity. There was a lot of inflow of capital during ZIRP leading to a lot of buildout, but with the end of ZIRP the low return on capital in their business was going to kill them. But then the AI bubble came along and everything within three degrees of bacon went to the moon.


reminder that according to Facebook's own analysis 10% of their 2024 revenue comes from scams and banned products

All tech problems are actually people problems.

once the Csuite builds their own dashboards, they quickly decide what they actually need versus what is a nice to have.


And I wonder if they will discover that in order to interpret those numbers in a lot of cases they will need to bring in their direct reports to contextualise them.

If corporate decisions could be made purely from the data recorded then you don't need people to make those decisions. The reason you often do is that a lot of the critical information for decision making is brought in to the meeting out-of-band in people's heads.


the initial rlhf training evaluation was done by kenyans specifically

it's easy to scan for publicly known services, really difficult to understand if a random string that says key somewhere is actually a random internal api key

which is why a lot of services now prefix they keys with a fixed string like pat_, sk_,

much harder to train longer context inputs

Everyone trains on queries from other models, it's called distillation

Well, given AI content cannot be copyrighted, haters can hate hate hate hate hate...

That's a little disingenuous. If i buy a printer and use the ink in the cartridge to reverse engineer a beautiful red, have i stolen something from the printer manufacturer? Especially if they lose business because they no longer have what distinguished them?

Clean room design is not new (or illegal), but it's always been a form of stealing


what's wrong with the current ide tools?

I feel like that's how you get Microsoft where each division has a gun pointed at the other division

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: