Alex the Parrot

[gusl]

Was Alex the Parrot a modern-day Clever Hans? Did Pepperberg try to exclude the possibility that Alex was just reading her?
These videos, if legit, are pretty very impressive (and cute).

How much can you do with just operant conditioning? What intelligence tests would you demand, to test their flexible uses of representations? Theory of mind?

Comments

[isomorphisms.livejournal.com]

I have a book about Alex, and I remember a few articles on the subject saying that when people other than Pepperberg tested him, he still performed far better than random.

I had a parrot years back, who said a few words. He knew the contexts, too: for instance, when the phone rang, he'd say "hello".

[gustavolacerda.livejournal.com]

Well, having other people would make it harder for him to read the answers from the human's face... but it's still quite plausible.

[isomorphisms.livejournal.com]

Sure, but how much of our own ability to answer questions and such comes from a learned ability to identify patterns, read other people, and such? I don't think that we should be downplaying the abilities of a parrot for using the same tricks humans employ daily. I am reminded of a friend who made it to the national level of some chemistry competition, not by studying chemistry, but by studying how multiple choice tests were constructed.

[gustavolacerda.livejournal.com]

Fair enough, but as an educator, you should be able to answer your question. You must be familiar with issues about testing and transfer, the holy grail of transfer...
As an AI person, I'm very fond of transfer (i.e. flexible representations) as a criterion for intelligence.

[zarex.livejournal.com]

I used to visit Alex when I got my coffee in the mornings at the Lab. I'd give him bits of nuts for correctly identifying the color of various objects. He was pretty convincing to me, and it was just us.