Your Lying Eyes

Dedicated to uncovering the truth that stands naked before your lying eyes.

E-mail Me

Twitter: yourlyingeyes

30 November 2008

Waiting for those Domestic Bots? Keep Waiting

An article in Science Times last week warns us that modern science is in no hurry to bring us the next generation of domestic time-saving miracles.
...the attempt to build thinking machines that can perceive the world around them and then act on that awareness. Researchers are far, far from being able to design a Rosie Jetson or a Data, or even a Diaper Data. You can ask a human toddler to bring you the red ball from behind the sofa, and the toddler will comply. Ask a machine to perform the same seemingly ho-hum task? “We’re not even close,” said Seth Teller of M.I.T.
Well, sure, most of us never consider how difficult it is to make out distinctly all the things we see - it's a months-long training program we begin from the moment of birth - and something our brains are likely hard-wired to perform. You couldn't just start doing it from scratch today. (Like in the gospels, when Jesus cures the blind man and asks him how things look. "I see men like trees walking" was his unsatisfied response. Jesus made a few adjustments and asked "There, now, how's that?" "Perfect.")

The great achievements of the 20th century in household technology largely served to liberate us (and mostly women) from the enormous burdens of keeping a clean house and tidy yard (an effort considered essential to the health of the family). Electrification of homes and advances in electric motors eliminated some very difficult work: beating rugs, bending over sinks scrubbing clothes; hanging clothes on clothes lines; washing dishes after meals; and, last but not least, getting her hair in shape for a night out. Men benefited from the development of power tools, in particular the gas-powered lawn mower. And the automobile made everyone's lives unimaginitively richer. And then there was the air-conditioner, which allowed us to retire to sunny climes otherwise uninhabitable.

So why have things stalled - is it really due to limitations in optical recognition technology? Is what we have the best we can do? Thanksgiving weekend in the Northeast is a ripe opportunity for imagining further advancements in domestic automation. I put up Christmas lights on Saturday - nothing too elaborate, but I needed to maneuver an unwieldy ladder (a redundancy) at odd angles to get my strings of soon-to be-outlawed mini-incandescent bulbs mounted in just the right spots. I don't like climbing ladders, and for good reason - they are a major source of injury. Wouldn't it be a lot safer to ascend in a cherry-picker or other hoisting device? Here's a website marketing various industrial hoisting devices that surely could be modified for home use. There's no technological limitation - the only problem is that we are not a wealthy enough society to have a cherry-picker in every garage. Had we continued to grow 5.5% each year as we did in the sixties, we might be wealthy enough.

Leaves are another source of wasted hours in the northeast. Paying people to do the job for you is one approach (particularly with cheap immigrant help), but not a sign of an advancing society. Why aren't there machines that can be operated remotely, covering one's entire property, can sense the difference between a football and leaves, rejecting the former while pulverizing the latter (we know such technology exists from watching "How It's Made"), and automatically spitting out fully bagged piles of powdered leaves along the way? This is not technologically insurmountable - it's just too expensive for the average household, and thus not worth developing. But again, if we were a wealthier society,

How about laundry? The washing machine and dryer were great inventions, but the long term effect seems to have been to allow households to own many more individual items of clothing, so that now laundry seems as big a job as it ever was. About the biggest innovation in household laundry has been to place laundry rooms on the bedroom level in newer homes. (There's also the ubiquity of dry cleaning thanks to immigrants willing to work dirt cheap.) How do we not have continuous washing systems, where one basically drops a piece of clothing in a slot, chooses a cleaning method, and walks away while the machine cleans, dries, and folds the article, depositing it in a neat pile? If a 3-inch camera can find people's faces in a viewfinder, how difficult is it for a machine to differentiate underwear from socks and shirts from pants?

I'm sure many of you have better ideas. The fact that our work and time saving technologies have not advanced beyond 1970 in any area of our lives outside of communications and data processing is testament to how little we have grown economically. We just haven't gotten appreciably richer since then.

5 Comments:

Anonymous Anonymous said...

Work on automation has progressed rather nicely in Japan where immigration is restricted. Labor in the USA is still cheap, wages are stagnant, and the reward for investments in automation is therefore small. A huge portion of GDP in the USA now goes into entertainment in one form or another rather than productive work.

December 01, 2008 9:41 AM  
Anonymous Anonymous said...

Portrayal of robot labor more prevalent than actual robot labor?
http://blog.wired.com/gadgets/2008/11/actors-robots-t.html

December 01, 2008 11:53 AM  
Blogger ziel said...

I'm sure cheap immigrant labor has a lot to do with it, but the fact is our growth has been much flatter of late.

December 01, 2008 11:29 PM  
Blogger Steve Sailer said...

Robert Heinlein's 1957 novel "The Door into Summer" about a man who invents the first household chores robot in 1970 (and then returns in 2000 to check up on how its working out) is interesting.

Heinlein got the hardware problems pretty much right -- his robot isn't humanoid at all. The inventor starts with a powered wheelchair, and then attaches mechanical arms to it. (It doesn't climb stairs -- you can buy a second one for your upstairs room.) What Heinlein totally underestimated are the software problems. For example, the inventor spends an evening programming it to wash dishes in the sink. Well, washing dishes in a sink turns out to be terribly, terribly difficult to program, so instead of a mobile jack-of-all-trades household helper, we've gone to a fixed location one-purpose dishwaser machine instead.

But, I think it would make sense for some corporations to work with home builders on prototyping the subdivisions of the future. For example, a Roomba-style lawnmower could work if you embedded those things under the edge of the lawn that are used to trigger electric dog collars to keep your dog on the property.

December 03, 2008 3:37 PM  
Blogger Steve Sailer said...

Another thing a subdivision of the future could have would be refrigerated storage accessible from the outside for delivery of groceries when nobody is at home.

December 04, 2008 4:04 AM  

Post a Comment

<< Home