News for hackers. Stuff that matters.

SeanaldMcDnld

Re:Genie 3: A new frontier for world models on Tuesday August 05, 2025 @12:22PM (#44800094)
Consistent output and spatial coherence across each eye, maybe a couple years? But meeting head tracking accuracy and latency requirements, I’d bet decades. There’s no way any of this tech reduces end to end latency to acceptable levels, without a massive change in hardware. We’ll probably see someone use reprojection techniques in a year or so and claim they’ve done it. But true generated pixels straight to the headset based on head tracking, is so so far away.
Re:John Carmack at Upper Bound 2025 (slides and notes) on Friday May 23, 2025 @09:52AM (#44072901)
Yeah in that way this demo seemed gimmicky like he acknowledged. He said in the past he would almost count people out if they weren’t training RL in a virtual environment. I agree, still happy he’s staying on the path of online continual learning though

Ask HN: What's the state of online RL with no pretraining

What's the current state of online continuous reinforcement learning? And where can I read more about it. I've seen a lot of research papers within the last few years on training RL models with offline data, or at most, episodic training. From what I've gathered, It's just too difficult to train agents online in complex environments because the policy conditions the training...

( Read More... | no comments | 2 points | on HN )