News for hackers. Stuff that matters.

baegi

Re:Qwen 3.8 27B available on Cerebras at 1500 tokens/s on Thursday September 03, 2026 @04:34PM (#49556527)
Not sure how the rate limiting works, but it's 1.5k TPS, not 15k, so you could run it for 100s/min, which seems good enough to me
Re:SpaceX Starship Flight 13 livestream [video] on Sunday July 26, 2026 @12:35PM (#49059773)
They planned on catching the booster on flight 6 as well, but aborted due to some issue with the launch/catch tower
Re:"We are hiring professional 'vibe coders.'" on Tuesday February 10, 2026 @02:18PM (#46965327)
... but not as coders. Funny in a way
Re:Python's new t-strings on Monday April 21, 2025 @05:19AM (#43749873)
except if get() can also accept a raw string, which is likely
From my European pont of view, I think the definitions of left and right have shifted a lot.

Sure, in Europe left and right may be more closely matched in academia, but most "right"-leaning Europeans would not be anywhere near the "right" in US-terms, so your argument is comparing very different things

Sure, if they also overpaid for their phone, they should enjoy the privilege of airdropping to gp. Just not from cheap devices please!
M136279841
Eh, I don't see "better than Elon's burning pile of garbage" as the greatest endorsement.
Re:How 'Factorio' Seduced Silicon Valley on Sunday October 27, 2024 @07:34AM (#41961739)
Pretty different to Factorio, but Human Resource Machine is somewhat relevant and very enjoyable
But wouldn't the model then also learn to make reasoning mistakes in the first place, where in some cases those mistakes could have been avoided by not training the model on incorrect reasoning?

Of course if all mistakes are corrected before the final output tokens this is fine, but I could see this method introducing new errors altogether.

I have seen so many claims like this one, yet never any evidence of this sort of thing actually happening. I tend to believe this is a combination of confirmation bias and "side-channel" information like googling related things, or pattern recognition on the ad server's ML side (e.g. if you visit the amazon toilet paper section once a month, and after almost a month you start talking about toilet paper, getting served toilet paper ads might seem suspicious, but it really isn't).

Is there any actual information about this somewhere? With the insane breach of privacy policies and laws this would be, I would think many researchers would have looked into this by now.

Those articles pop up every now and then, and it usually turns out to be the cargo trunk of the Dragon spacecraft, which gets jettisoned before the craft performs its reentry maneuver. The trunk comes down usually a few months later. It survives reentry due to its low density carbon construction.

Starlink satellites, of which there are thousands in orbit, are designed to burn up on reentry.

Re:World in a Box on Monday May 27, 2024 @07:36AM (#40489842)
To be fair, your comment was in no way interesting. You announced that you closed the tab without reading the article, providing no value to anyone. I don't see why you're surprised about a lack of votes on your comment.
Re:Antarctic Ice Shelf Spawns Iceberg A-83 on Monday May 27, 2024 @04:47AM (#40488921)
I'm curious as to why you're looking forward to those events. Would you mind elaborating?
Re:Meta does everything OpenAI should be on Thursday April 25, 2024 @03:52AM (#40154681)
Can you elaborate?
The 3090 has 24 gigs as well and is available for a very decent price on the second hand market
Re:Peregrine moon lander suffers anomaly after launch on Monday January 08, 2024 @04:36PM (#38918532)
"Just before entering a known period of communication outage, the team developed and executed an improvised maneuver to reorient the solar panels toward the Sun. Shortly after this maneuver, the spaceraft entered an expected period of communication loss."

Imagine the pressure when clicking" send" on that maneuver patch