The press is excited about Google’s purchase of robotics companies. For example, Popular Mechanics writes,
Car factories are stocked with powerful, precise robotic arms that assemble cars with uncanny speed and delicacy. Bringing that to our world—to our streets, nursing homes, warehouses—is the next big step in robotics. And Google bought its way into this future.
The interesting conflict in robotics is between those who want to build general-purpose, humanoid robots and those who want to build robots optimized for specific tasks. I lean more toward the latter. For example, I expect to wait a long time (perhaps forever) for a robot that can do all-purpose house cleaning. It is easier for me to picture a hotel using a toilet-cleaning robot, a shower-cleaning robot, a bed-making robot, etc., as such devices are invented.
An inquiring mind asks, will Google develop Robotic Soldiers, for land wars throughout the world?
More likely remotely piloted vehicles for bringing firepower into contact with the target list and some autonomous systems for surveillance and logistics (including bringing fuel and ammo to the firepower platforms). It’ll be like playing one of those complex war simulator games.
There might be a few situations where you need an suicide-assassin bot, maybe a mouse-bot or wasp-bot, something very small with maybe a hand-grenade at its core, that can scurry or fly quietly through very tight spaces and preferably in the dark, and get right up next to the target and then blow up.
It seems to be that assassin-bot technology is perfectly feasible right now, and it is only going to get ‘better’ with time. The Secret Service has no tool to fight a large swarm of explosive wasp-bots full of poisoned shrapnel. The government is going to need a lot more surveillance to prevent such attacks, if it’s even possible. One successful attack by assassin bot is going to be like another 9/11 in terms of panic and new authorities.
The future is an exciting, but also very scary place because of all these robots. Manual is Over!
So finally the Internet meets at meets robotics. Sounds a lot like skynet is around the corner. However I believe that an all purpose house maid is as far away as Tyler Cowen fears about robots replacing men. Yes simple tasks will get automated when the costs of manual labor exceeds the costs of automation. But we are on the limits what is possible with simple algorithms in most robotic applications that include multiple choice actions. If you ever called a service Hotline with an automated Q and a process you see it clearly. Another example is Google now or Siri, both are OK but have sub par performance IMHO. Most times they misunderstand stuff and they have problems distinguishing languages.
The same goes for robots. As Mr Kling said, we will see robots targeted at specific problems, but vacuuming with a roomba f. E.
We tend to call it “robotics” if it either looks sort of humanoid (welding robots) or does something we consider amazing now.
We tend to call it a dishwasher, clothes dryer, roomba, or automatic sprinkler with rain detector when it does something we are used to. Even if it would have been mind boggling in the relatively recent past.
And while there were surely images of robot controlled engines, since in the real world the automation that controls engine settings in response to your throttle input is invisible to you (unless there is something very wrong) – that particular branch of automation and robotics has now simply disappeared from view.
Arnold is right – it will be special machines, and they will become mundane, and we will not think of the Infirm Person Bathing Machine as a robot, we’ll call it by some brand name and shop for it like we do dish washers.