Flying to CES tomorrow, and the code is finally working :) Ready to demo CORE R3. It can now autonomously create new expressions, new movements, new dialog (even in Hindi), completely different from the original captured data.