croemer

24 ms·

451.
by croemer 1y ago
He has maybe mentioned to someone that he'll write a paper on this topic
452.
by croemer 1y ago
Rehabilitation is great. But you might have the wrong idea about the crime. It wasn't just addiction. He had enough U-47700 for 30,000 trips (30 thousand). See https://news.ycombinator.com/item?id=44291172
453.
by croemer 1y ago
I'm not judging anything at all. What part of my comment makes you think I judge the worthiness? I just decided to share what the crime was since OP left it out. To make it unambiguous I added a prefix: "Great story, I wish this i
454.
by croemer 1y ago
3,000 not 300 if my maths are correct (and lethal dose)
455.
by croemer 1y ago
Great story, I wish this inspired more prisons around the world to follow suit. For those who don't want to hit Google, the conviction was for possessing 30g of a synthetic opioid "U-47700". A normal dose is ~1mg, 10mg can be
456.
by croemer 1y ago
Turns out there was no notice. So it's Slack pulling the rug in an aggressive way: https://github.com/kubernetes/community/issues/8490#issuecom...
457.
by croemer 1y ago
Github issue for discussion: https://github.com/kubernetes/community/issues/8490
458.
by croemer 1y ago
4 days notice period to archive everything? Not a good look for Slack and/or the k8s working group (if they had known earlier already).
459.
by croemer 1y ago
No money gained because they move off from Slack.
460.
by croemer 1y ago
Same. And annoyingly you're not allowed to reuse old passwords, so you have to keep inventing (and remembering) new ones.
461.
by croemer 1y ago
Back up.
462.
O3-Pro(simonwillison.net)
2 points by croemer 1y ago|0 comments
463.
by croemer 1y ago
Dupe of https://news.ycombinator.com/item?id=44234290 and https://news.ycombinator.com/item?id=44234552 and https://news.ycombinator.com/item?id=44255352
464.
by croemer 1y ago
Dupe of https://news.ycombinator.com/item?id=44234290 and https://news.ycombinator.com/item?id=44234552
465.
by croemer 1y ago
There are US taxpayers and/or voters who don't live in the US
466.
by croemer 1y ago
Thanks! I could not find it
467.
by croemer 1y ago
Nature paper: https://www.nature.com/articles/s41586-025-09005-y
468.
by croemer 1y ago
Yes, my health relies on it. I sideload a glucose monitor app that's not available through Playstore (it's FOSS and health is a tricky area with liability). It's a fantastic app and the ability to sideload it is a major reaso
469.
by croemer 1y ago
Original study that's discussed in the Science News link: https://royalsocietypublishing.org/doi/10.1098/rsbl.2025.001...
470.
by croemer 1y ago
Update: It seems the author read this and fixed it. Thanks!
471.
by croemer 1y ago
Don't call "Nature Communications" "Nature". The prestige is totally different. Also, altmetrics aren't that relevant, maybe if you want to measure public hype.
472.
by croemer 1y ago
o3 totally nailed it first shot. Hiroshima since 2023. Provides authoritative source (Venetian city press release): https://chatgpt.com/share/683f638a-3ce0-8005-91d6-3eb1df9f19...
473.
by croemer 1y ago
*backing itself
474.
by croemer 1y ago
I love the channel. Carl is such a good YouTuber, full of insight, without doing attention manipulation (sure the titles and intros are good, but it's not over the top like most other channels).
475.
Internet of Bugs Channel(youtube.com)
1 points by croemer 1y ago|0 comments
476.
Vibe coders inside you (Programmers are also human) [video](youtube.com)
1 points by croemer 1y ago|0 comments
477.
by croemer 1y ago
Would be good to see the receipts of the apology.
478.
by croemer 1y ago
People have similar biases, it's just a lot easier to test an LLM for biases than it is to test humans so we are more aware of LLM biases.
479.
by croemer 1y ago
> Only humans can decrease or resist complexity. Proof needed
480.
by croemer 1y ago
> LLMs as they currently exist cannot master a theory, design, or mental construct because they don't remember beyond their context window. That's not categorically true: if a theory/design fits in their context window it&

More ›