Modax wrote:Concerning the human-like robots of the future: do you think its more likely that said robots will be powered by
artificial muscles or hundreds of micro-sized electric motors?
Current artificial muscle technology has rather limited power-to-weight. Geared motors are better but rather noisy. AFAIK the best near-term (non-nanotech) solution is linear electromagnetic actuators, which require more sophisticated control systems but have even better power to weight and are silent. Unfortunately they're expensive ATM, but there's no fundamental reason for that, with mass production they should be cheaper than rotary motors. If the goal is exact human replication then ultimately you'd want an artificial muscle fibre; as your article notes there are various (currently theoretical) ways of making that with nanotechnology that are wildly superior to the human version.
LordOskuro wrote:On a partially less spammy note, hopefully they'll get on to the placing human brain in robot/copying human mind into CPU issue too.
You do realise that is an almost completely independent problem being studied by a completely different group of researchers?
Actually strangely enough there seems to be relatively general AI research coming out of Japan - they have a few low-key brain simulation efforts, but nearly all the budget and talent seems to be going on applied robotics. This may be due to the embarassing failure of the 'fifth generation computing' project in the 1980s (which was a very large government funded general AI research program).