Fascinating video about some Japanese researchers who took the usual chimpanzee pattern recognition tests to a slightly higher level. Watch what happens about halfway through this video. Jaw-dropping, indeed!

 

http://www.youtube.com/watch?v=khe9bXT-OHE

 

The chimpanzees were never educated on the meaning of these symbols, so it seems fairly obvious that they're still just rote-learning the correct sequence, but isn't it fascinating that they are able to evaluate sequential order to this degree?

 

Also the short-term memory angle is interesting. I wonder if the reporter here has a point in suggesting that if humans spent that many hours on it they'd be able to do the same, but I think he also has a point in saying that evolution would surely suggest that chimps might well be able to exceed our capabilities in certain areas. Why not?

Planet of the Apes anyone?

I caught this on the news this morning, the reporter stated that children have displayed similar aptitude in short-term memory tests, but this generally fades with age. What's more interesting is a statement by Dr Lisa Parr (Emory University in Atlanta), from the BBC website...

 

"These studies tell us that elaborate short-term memory skills may have had a much more salient function in early humans than is present in modern humans, perhaps due to our increasing reliance on language-based memory skills."

 

More here...

http://news.bbc.co.uk/2/hi/science/nature/7124156.stm

Yet more evidence zoologists are baked out of their minds more often than not. It's no coincidence Planet of the Apes was all the rage in the late-60s, early-70s.

Strictly speaking, they proved that the chimps are better at this task than university students rather than humans.

Would this fall under the category of sensory memory? Some of the display times were pretty short. It would make sense that chimps have very good sensory memory.

Not to take anything away from the chimps, but they had to be taught to respond this way, and hence have practiced countless hours. The humans used for comparison likely didn't devote such time at practice, nor were they likely reinforced so consistently for the behavior.

How about the entropy of the information (in a chimp's brain)? they 'forget' how to do this stuff a lot more quickly than we tend to forget things. If the entropy is higher (they're 'faster' at this particular intelligence test), does it dissipate more quickly too?

Just a.. thought...

One thing I've been wondering about this is whether the chimps are practicing a simple pattern-elimination algorithm, or if they're actually learning the relative values of the 0-9 digits. The latter would suggest that learning is taking place at a level higher than just rote learning. If not, the sheer number of combinations they would have to try, and the number of patterns they would have to memorize, would quickly become really, really huge, wouldn't it?

 

But I'm just guessing; it might be interesting to calculate the number of combinations and amount of time it would take to test them all using a linear sort. Obviously the chimp will always hit on the right answer before finishing the test, but I can't help but wonder if they're doing a little better (or perhaps quite a lot better) than O(n) here.

The research I've read implies that they are not only learning successful completion of the task, but also context and relative values. I don't have any citations at hand right now, but it likely wouldn't be hard to find some.

Strictly speaking, they proved that the chimps are better at this task than university students rather than humans.

It would be funny if another university did this experiment and their "university students" were better than the chimps.