How do humans learn dynamic physical skills? How much is built in through biology? What are the built-in biases that help us learn some things easily? How can we make robots behave more like humans?
I'll talk about how human motor skills work from a robotics point of view and how we've applied some of those lessons to real robots we've built.
Slides:
There is no downloadable version of the slides for this talk available at this time.
About the speaker:
![[speaker photo]](090401-blackwell.jpg) |
Blackwell is the founder and CEO of Anybots, a Silicon Valley startup developing next-generation teleoperated robots. He is also a partner in Y Combinator, a new kind of venture firm specializing in early stage startups. Previously (1995-2001) he developed some of the first e-commerce software at Viaweb (acquired by Yahoo).
As a grad student at Harvard (1993-1998) he worked on randomized network protocols, randomized compiler optimizations, and other random things. He is published in the proceedings of ACM SIGCOMM, IEEE Infocom, and Usenix. He has also done a number of recreational hacking projects such as a balancing scooter, an electric unicycle, and an ultra-realistic Apple ][ mode for xscreensaver. |
Contact information:
Trevor Blackwell
320 Pioneer Way
Mountain View CA 94041
6507767870
6507452487
tlb@anybots.com