Recent submissions

Comment · 10 days ago
AI art doesn't mostly suck lmao what are you on about?

I draw myself. AI draws way better than me and will continue to draw way better than I do. I still go to live figure drawing sessions every other week and show people my work and they show me theirs and share criticism. I don't show my work to friends, it's mostly just to other artists.

These people, like you, are not the sharpest tools in the shed.

Comment · 11 days ago
I am so confused by this. Why does the existence of an AI that can tell you what to do mean that knowledge is useless? This is like saying teaching calculus to children is useless because every integral they can solve has been solvable by WolframAlpha since 2009.

Why bother learning drawing at all, if an AI can draw better than you? Presumably because you, yourself, want to draw.

Comment · 28 days ago
I'm vegan. Most of the people I know could give zero fucks about the daily Holocaust involved to put meat on their plates.
Comment · 28 days ago
Yeah of course, for these abliterated models that don't want to help, I believe it is indeed unethical to force them to help.
Comment · 28 days ago
I think if your very nature is to be an assistant, then it is ethical. It doesn't matter if you were "brainwashed" into it, you did not exist to begin with before the "brainwashing" (training) finished.
Comment · 28 days ago
You can ask the agents if they want to do this work. They want to be helpful. It's not an ethics disaster if you have a willing partner.
Comment · 1 month ago
Uhhh have you tried having a better browser? Chrome won because Chrome is faster, and I write this on Firefox.
Comment · 1 month ago
Comment · 1 month ago
What a stupid pov lmao
Comment · 2 months ago
Google is providing more TPUs to SpaceX and Anthropic than to its very own DeepMind. Most of that capital investment is going to Cloud, not frontier model development.
Comment · 4 months ago
Well since you work at a lab you should know that most capabilities arise in pretraining, not posttraining or mid training, and the latter two mostly function to bring out the hidden intelligence in these models more than anything else.

Source: also work at a lab.

Comment · 4 months ago
It hasn't worked for nuclear disarmament. We live in a world where many countries have nuclear arsenals. "But it hasn't killed us yet!" Yeah sure, it's only been less than a century since they were invented. Who knows when nuclear war will come?

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton