The first time I watched the writer and hacker Josua Klein’s crow vending machine TED talk as a college undergrad, I was floored. It was my first exposure to Betty, to the tool making capabilities of some crow species, and to the idea you could potentially train wild crows. The purported success of the vending machine filled me with ideas. I used clips from the talk for a variety of public outreach presentations and they were always met with the same kind of GTFO amazement that I love watching people experience as they learn about crows.
As I moved on to graduate school, however, and was fully immersed in the scientific community of crow nerds, I started to hear rumblings that gave me pause. Rumblings that suggested the vending machine wasn’t all that it was cracked up to be and, in fact, had not worked as it was implied in that TED talk. Since I’ve never worked personally with Klein, I’ll let my fellow crow scientists speak for themselves on the issue. You can find one of the graduate students he worked with relating her experience during a reddit AMA here, and as well as the correction that the New York Times Magazine was forced to run after publishing an article on Klein’s effort with the vending machine. If you don’t want to read them, suffice it to say the main point is that Klein gave people the impression that it had been tested (successfully even) on zoo and wild crows when it hadn’t.
Leading the public to believe that we’ve arrived at conclusions when we haven’t is the stuff of stress dreams for scientists, and it’s why the peer review process is the foundation of good scientific practice. By taking “results” that were only in the early stages of being tested and bringing them to the attention of the public without permission or support from the scientists he was working with, Klein burned his bridge to the folks who had offered to help him test the idea, and any other crow scientist he might approach next. Which brings me to the recent article I read titled “This Machine Teaches Wild Crows to Bring You Coins for Peanuts.”
No, it doesn’t. It might, but probably not. No one has been able to train wild crows to bring specific items in exchange for food, the website selling the machine even points this out. Gabi Mann did not intentionally train her crows to bring her things. They did this of their own volition which is why her collection is as diverse, unique, and beautiful as it is.
The suggestion that this machine could train crows to bring you quarters holds about as much water for me as saying you could use a dog whistle to train wild wolves to roll over on command. The reason that the machine worked on captive birds in the Brooklyn apartment where it was originally tested is that, in captivity, you have a certain amount of leverage over an animal. You can motivate it with food or treats or affection. The chances that a wild crow would go to the effort of looking for coins when it could simply skip that step and look for other food seems insurmountable.
All that being said should you turn your nose up at The Crow Box if the idea intrigues you? No, go for it! Maybe yours will be the mind to figure out how to motive wild birds to participate. Or, perhaps you don’t care if it works or not, you’re just in it for a new experience or the joy of trying. Trying and failing is part of discovery and I see no reason people should wash their hands of it if it sounds like fun. Plus, even if it doesn’t work, you may end up learning different, but just as amazing things about these birds. Just don’t hold it against the crows if they decide it’s simply not worth the trouble and leave it to you to go collect the quarters you lost buying The Crow Box.
8 responses to “A scientist’s thoughts on The Crow Box”
Pingback: Top ten - Ocasapiens - Blog - Repubblica.it
I think you’re right to be skeptical of the Crow Box, although I am sure that a human could train them individually to go out and find coins. (Quarters, probably not; they’re rare on the ground and the crows would give up. Pennies and dimes, yes.) Perhaps once a human has done it, someone could figure out an automated training process, but that’s *much* harder than having a trainer in the loop who can keep adjusting the training and fit it to the individual crow.
Probably a much better plan is to train a couple wild crows by hand, and perhaps other crows would pick up the trick by observation.
« The suggestion that this machine could train crows to bring you quarters holds about as much water for me as saying you could use a dog whistle to train wild wolves to roll over on command. »
This I have to take issue with. People have used reinforcement training to train wild animals to do things, including fish and zoo animals, so I don’t see why not!
Hi Tim, give me an example of people training free ranging wild animals (which is what I meant by wild wolves, as opposed to captive wolves) to DO things (as opposed to using negative reenforcement to train them NOT to do things). I don’t say that to be combative, if you know of an example I would love to hear it, as it might change my perspective on how possible this is.
I know this is an old post but with food and working with their inclinations wild animals are pretty quick to catch on.
Again Jennifer, can you (or anyone) provide an example? An example of a wild, free ranging animal that’s been trained to actively do something? I’m not talking about shaping their arrival to a predictable food source, or to use a food dispenser, but an actual two-step task. It’s that first step, the one that’s no directly relevant to obtaining food that seem insurmountable, but that I am hoping someone can provide an example of.
Just a point of clarification, the Crowbox webite is NOT selling anything, they have the plans and software available for free. They do provide a complete parts list and assemble videos. They also link to a commercial site that sells the materials and will laser cut it for you. . You need to buy the materials and electronics and assemble everything yourself.
Found this blog today, thanks for writing it, love to watch the crows in my yard.
I thought this YouTube video may interest you — he appears to have taught magpies to deposit bottle caps in exchange for food:
He has other similar videos on his channel; however, this video above shows the progression of his “bottle-caps-for-food” device and how he trained the magpies.
Pingback: Butts for nuts: can crows do our dirty work (and should they)? |