Keep on to blur preview images; turn off to show them clearly
Okay fine. For those of who who somehow think this is a skill issue despite me being able to build crazy stuff without LLMs: if someone I respect replies to this with the exact setup I should use, I will spend a day live on stream attempting a mid-size project exactly as described in good faith. If it is a net positive, yay, I update my workflow and get more stuff done. If not, I will deepen my conviction that these damn things are mostly making everyone lazier and stupider. I will have RL ~solved in ~2 years with or without LLMs either way.
I build sane open-source RL tools. MIT PhD, creator of Neural MMO and founder of PufferAI. DM for business: non-LLM sim engineering, RL R&D, infra & support.
The reason people think the models are fake and don't work is because the models are fake and don't work. You hear about all these breakthroughs and then go to use the latest model just to find out that it's only superhuman at lying to you.
I build sane open-source RL tools. MIT PhD, creator of Neural MMO and founder of PufferAI. DM for business: non-LLM sim engineering, RL R&D, infra & support.
Want game agents that you can run on a potato? No VLMs needed. We have tiny RL models you can demo on https://t.co/T2xjeFieYt. Train most of them in seconds on your laptop!
I build sane open-source RL tools. MIT PhD, creator of Neural MMO and founder of PufferAI. DM for business: non-LLM sim engineering, RL R&D, infra & support.
Hello - I am alive. I took a vacation. A real one, for once! I'm working on infra off-stream for a few days and will be back streaming shortly after Christmas. The way forward for RL is clear, but the infra required will make PufferLib 3.0 look like a toy!
I build sane open-source RL tools. MIT PhD, creator of Neural MMO and founder of PufferAI. DM for business: non-LLM sim engineering, RL R&D, infra & support.
Hey, that's my first paper! I believe I was 19 at the time. I read a few papers, put two that I thought were complimentary together, got SOTA on a small task, and did a little bit of interesting analysis
I build sane open-source RL tools. MIT PhD, creator of Neural MMO and founder of PufferAI. DM for business: non-LLM sim engineering, RL R&D, infra & support.
Taking a vacation. Back to finish Puffer 4.0 soon! Pulled 405 raw on the last stream for now. 172 lbs 5'9 https://t.co/AqYFK60yTe
I build sane open-source RL tools. MIT PhD, creator of Neural MMO and founder of PufferAI. DM for business: non-LLM sim engineering, RL R&D, infra & support.