Recent submissions

Comment · 3 days ago
"Cyber benchmarks, while achieving an average refusal rate of 96.6% on malicious cyber requests, the highest we measure among open models."

I feel like stating the obvious here and hence like I'm missing something. Where is the statistic about not refusing the non malicious requests? I can achieve the same performance with refusing when random(1000) < 967.

Comment · 4 days ago
Note that DST doesn't change that. It makes the already long days feel longer, not the short ones. Or you could view it as your mornings being brighter in winter perhaps. But yeah, as such it never made much sense to me, though maybe my view is too much mid latitude driven.
Comment · 4 days ago
Of course not, but that's a separate problem that I didn't treat. It was about the ability to make that choice, I don't think that that my choice should be respected is a controversial statement worth debating.
Comment · 4 days ago
Choose whether you want logs when creating your instance? Good practice.
Comment · 5 days ago
Every system gets gamed, and so is the current one. The goal is that the people can more easily stay informed. Even with mixed authors problem you highlight would be still an improvement.
Comment · 5 days ago
The difference being that the M-series machines do not have good prospects in terms of running Linux well. As I would personally not want to run Windows for sure, this matters a lot.
Comment · 10 days ago
The key was enough of a statement of TLS's place in their business. No new information here.
Comment · 14 days ago
Honestly, the biggest discovery to me is that the game was created later than I expected. 1984 in the video title would put it in the same year as Duck Hunt; Wikipedia seems to suggest 1985 was the actual year it was published, which is the same year as Super Mario Bros (which itself came few years after Donkey Kong).
Comment · 14 days ago
True, I conflated decay with fission. So another source indeed!
Comment · 14 days ago
"Why have X when Y costs $N in tokens" is becoming a tired thought-terminating cliche.

It shouldn't be user's responsibility to have to get their binaries to work at all costs.

Comment · 15 days ago
Arguably nuclear fission in this case.
Comment · 18 days ago
That's my view. The end goal is to become Too Big To Fail, the means is stifling the competition and getting military funding.
Comment · 19 days ago
>I've learnt more since the advent of LLMs than I did in the 10 years prior. And it's not just me; that is most people's experience with LLMs

My own usage too. Yet I believe that your environment is biased towards ambitious adults aligned with your values, who want to learn things. I don't think it's representative of world as a whole, and I'm even further worried about the youth, being a particularly vulnerable group. Let's hope that teachers and parents of today and tomorrow will keep reinforcing the idea that knowing things and having skills is worthwhile.

Comment · 1 month ago
I would hardly call the requirement to understand the code you push "gatekeeping".
Comment · 1 month ago
Meaning, you wouldn't touch it with a ten feet pole?

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton