Hitachi Unveils Humanoid Robot

HunahpuMonkey writes "BBC reports that Hitachi has unveiled a humanoid robot, named Emiew, to compete with Honda's Asimo and Sony's Qrio robots. The robot has a vocabulary of about 100 words and could be trained for practical office and factory use. In addition, it is the fastest robot to date, moving 3.7 miles per hour on wheel feet which resemble the bottom half of a Segway scooter."

Comments

I've seen something like this.

[Sheetrock]

During a brief sojourn to Japan a couple of years back, there was a music store that had a robotics theme.

Very little stock was on hand, but you would select the music you wanted on one of the robots. It'd burn the audio CD, print up the liner, and assemble a shrink-wrapped product for a couple of yen more than one you'd get off the shelf, then dance around the room playing the biggest hit off the album.

The experience would only have been cooler if it could talk with you, although the sushi-dispensing robots did have a few stock phrases and voice recognition (you had to shout for them to hear you however).

Why robot research is wasteful

[Tablizer]

Japan is pouring billions into robotic software research in part because they don't allow much immigration and migrant workers, and thus want to develop robots to fill those niches instead.

However, rather than build an artificial brain, it appears more cost effective and closer to improve the bandwidth costs so that such bots can be controlled from low-wage nations. We don't need artificial intelligence because there are billions of idle human brains around the planet.

I suppose one could argue that remote-control servants could end up causing malice, but artificial alternives may do the same either because AI might go bizerk, or more likely because it is not good enough yet and will make stupid mistakes.

In short, remote-controll appears the more reachable goal at this stage. Bandwidth cost reduction does not appear to need the giant breakthrus that AI does.