I'm working on AI for robotics. My goal is a cheap robot that can do all your household chores.
Attention all students! I'm looking for people with the same goal. We have the robot, we have the compute, let's make it work! I care about practical results only, no theoretical advantages. Run it on the robot and show that it works!
Cheap robots are easier to experiment with. Everyone can have them and it’s not a huge loss if they break. If they are open source we can all improve them together and create something more versatile and robust.
We have AI that disproves decade-old conjectures. Surely we can have a robot that does the laundry. Recent developments have been very promising in that direction. It appears that you can simply scale up pre-training to significantly improve your models, similar to LLM progress in the GPT-3 era. Modern VLMs provide a strong basis for this training and progress is only speeding up.
