glaksmono

I noticed that some people are working on trying to close the gap of the Sim2Real, but most of the videos are on the physical stuff that are rigid. Sim2Real gap happens when the stuff are more unpredictable (i.e. cloth, rubber, etc)

Thoughts on this?

I don't think there's a use case that really "works" yet at the moment and the exploration is still undergoing

curious if anyone have seen an interesting use case of humanoid robotics that "works"

I found an article that says the following:

"CANN 6.0 claims to support one-click migration of CUDA code, and compatibility of PyTorch framework of around 95% according to some sources."

Source: https://pstaidecrypted.substack.com/p/gpu-games-heat-up-just-as-ai-diffusion#footnote-2-162340886

I'm not sure why, I find o1 is better than o3-mini for most of the questions I'm asking

Anyone seeing something similar?

I can't pin point down on exactly why

It's insane to see how much I can get done on my own with AI. It's a dream for the idea guy

I will only be hiring senior-principal guy who can leverage AI as part of their work

Which do you guys use, and why