Archive for January 25, 2010

Good Luck With That

How did I miss this?

Back in April of 2009, Wired.com reported that the United States Department of Defense wants to replace dog trainers with robots.  The Defense Advanced Research Projects Agency (DARPA) is funding research to “Develop and validate a portable device that automates the training of complex behaviors in animals without human intervention”

DARPA says they want robot dog trainers because:

Animal training currently requires long hours and the involvement of a human trainer.  The development of an automated mammalian training device would significantly reduce the need for human involvement.  In addition, it may enable the ability for remote on-site training in potentially limited access areas.  This device would also have the ability to better and more rapidly train an animal through the collection of performance metrics that indicate subject intelligence, capability, and progress.  Animal use is anticipated under this topic.

Of course – everybody knows that dog training requires nothing more than the rote implementation of a simple four quadrant operant conditioning algorithm. Even a robot can do it!

DARPA’s goal is to create an automated device that can train dogs to discriminate between objects, respond to verbal cues, retrieve objects, excel at tracking and more. Oh – and they want the doohickey to be cost-effective and portable too.

DARPA may have millions billions to spend, but I’m not planning for my retirement yet. If you’ve done enough dog training to accomplish much of anything, you know that it takes a lot more than well-timed rewards and punishments to train a dog. Teaching a dog how to do complex tasks reliably in the face of distractions is art, not algorithm and I think DARPA’s wasting taxpayer money.

Can DARPA build a computer that can read and correctly interpret canine body language? And can they make a robot that’s capable of using that information to communicate with dogs at a level that will allow it to train them to do complex tasks?

When I train a dog I don’t a complete a task, I enter into a relationship. A relationship built on trust and communication. A machine just can’t do that.

A computer probably has faster reflexes than I do, and it will probably work more hours for less pay – but it doesn’t have a soul. Dog training is an art and it takes years of mindful practice to do it well. When DARPA successfully builds a computer than can compose music or create sculpture, they may be ready to move on to dog training.

A machine won’t see the subtle shift in posture that tells me when a dog is confused and needs help. A computer can’t sense when to speed an exercise up, slow it down, make it simpler, add distractions – or just give the dog a heartfelt word of encouragement. I don’t think a robot will be able to tell how and when to transition between play and work to keep a dog motivated; or know exactly when to give it a break to process what it’s learned.

Most importantly, I don’t think that a creature that evolved to be a helpmate and companion to man will want to work for a machine. My dog doesn’t work for well-timed liver treats or tug toys – he works because he finds joy in the work – and in our relationship. No machine can replace that.

January 25, 2010 at 11:54 pm 8 comments


Because A Dog’s Mind is a Terrible Thing to Waste

RSS

Copyright notice

All original content on this Web site is copyright © on the date of publication by this author. All rights reserved except, of course, that others may quote from original content under the 'Fair Use' provisions of US copyright law.

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 285 other followers

Top Dog Blog
Featured in Alltop
ResearchBlogging.org

sitemeter

Add to Technorati Favorites
Dog Blogs - BlogCatalog Blog Directory
blogarama - the blog directory
Blog Directory
Blog Directory & Search engine
January 2010
M T W T F S S
 123
45678910
11121314151617
18192021222324
25262728293031