Recent submissions

Comment · 2 years ago
Docker
Comment · 2 years ago
Did they have to add a label that the goods may give them cancer?
Comment · 2 years ago
Humans try to replicate the human brain in software and are surprised it sometimes spits out dumb things a human could have said.
Comment · 2 years ago
An area worth exploring are IP cameras imho

1. They are everywhere and aren't going anywhere.. 2. Network infrastructure to ingest and analyze thousands of cameras producing video footage is very demanding.. 3. Low power and low latency scream asic to me

Comment · 2 years ago
Any other creative pursuit that involves some level of engineering.

Buy a 3D printer, start some woodworking/furniture stuff, learn to weld. Dig a pond and breed koi fish.

Comment · 2 years ago
Fair enough, at least the limits they proposed are among the best in the world. Credit where credit is due!
Comment · 2 years ago
"First ever" has been used very leniently here. Maybe it's the first ever for any state in America but the rest of the world definitely already has limits on it.
Comment · 2 years ago
One word: fuzzing.
Comment · 2 years ago
The best vegemite I ever had was from a 6 foor 4 muscular man in Brussels
Comment · 2 years ago
Does it support OpenVPN?
Comment · 2 years ago
Did they get consent?
Comment · 2 years ago
I got some FreeStyle Abotts for anyone who wants to hack away at them. Rumor has it Dexcom is cheaper in bulk than Abotts and they are losing hospitals as customers.
Comment · 2 years ago
I've tried out HerbGrind but it often gets stuck and reports an error E_STONED.
Comment · 2 years ago
Breaking News, BuzzFeed man can take a joke amd fires back.

I appreciate defining a clear hypothesis and the exploring an LLM using statistics. I feel like the analysis could benefit from prompts that contain neutral consequenses as well. You have given it clear positive rewards, clear negative ones and no reward. Neutral consequences may be a better baseline than no reward.

Comment · 2 years ago
TL;DR BuzzFeed man performs statistical analysis on LLM output, in an attempt to determine hidden internally encoded motives.

Next up, for a more clickbaity titel, BuzzFeed man pretends to be therapist to uncover LLM's dark secret.

I only wrote this snarky comment because 90% of the authors job is to evaluate the effectiveness of their clickbaity titles, or am I wrong?

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton