Hacker News Reader

Do we know much about the reliability of crow behavioral science (or whatever the field is called, if it has a name)?

Josh Klein, the fellow who popularized the ideas of crows being trained by a machine to retrieve things for reward, turned out to be extremely misleading[1] about the circumstances and reliability of his research.

Granted, Joshua Klein is not connected to the startup in the popular mechanics link, but interestingly, that startup does cite Klein as their source of inspiration[2] on their website.

1. http://www.nytimes.com/2008/12/14/magazine/14Ideas-section4B... (see editors note)

2. http://www.crowdedcities.com

How does it tell the difference between a crow it has seen before and a new crow? Would it just dispense free peanuts to all crows?

It dispenses in response to cigarettes, not crows (at least after training), so it wouldn't really matter I don't think.

> (at least after training)

That is the question.

The article described "training" as progressing from free peanuts to peanuts for butts.

Maybe after the initial training period they just rely on crows learning from each other

Amazing ... now we will have crow butlers picking up cigarette butts.

Not a problem in NYC by the way ;-)