I had a small back-room contribution to the DRC Finals that ran this weekend (final results can be found on the DRC website: http://www.theroboticschallenge.org/).
While a lot of progress has been made there still appears to be problems to overcome. The intent of the DARPA competition was to force the teams to build robots that could operate as autonomously as possible. To that end the communications between the robots and human controller was "degarded", which is the portion I was involved in. (I'm always causing people problems! At least now I get paid for it.) The teams had a 9600 bps bidirectional channel to work with at all times and an intermittent 300 Mbps unidirectional (from the robot) channel intended for sending of image data, if needed. Most seemed to need that data stream.
I think the DARPA challenge results are somewhat relevant to some uses of nanotechnology because the DARPA constraints are roughly analogous to those that nanobots would operate under (e.g. in nanomedicine.) Instead of having to balance themselves, though, the nanobots would have equally challenging problems staying on station, for example.
If I had to characterize my view of AI and robotics advances in a few words, I'd say capabilities are growing linearly, not exponentially, and there is nothing to indicate to me that advances are being parallelized to make the rate of advance exponential.
(I got a number of personal photos and videos - mostly redundant or inferior to what you can find elsewhere on the net so I'll only share one of this poor fellow, who took a fall - a common failure mode, it turns out):
Don’t have an account yet? Register yourself now and be a part of our community!