That's me in the corner…..

<

p>

<

p>One of the mind numbingly boring perks of traveling so much for work is that I get to read A LOT. The book I am currently reading is called On Intelligence by Jeff Hawkins. You might remember him for starting such companies as Palm and Handspring.

<

p>However, this book is not really about electronics, and it’s almost not even about computers at all. It turns out that Jeff’s real passion is the brain and learning how it works. His objective is to create a new approach to A.I. One that is much different than the current direction of the field of A.I.

<

p>One of the interesting points he makes early on is how the field of A.I. is going down the wrong path by focusing on the behavioral aspects of intelligence. Simple regurgitation of facts or the processing of a set of inputs to produce some sort of output cannot be considered real intelligence. This is the path most of A.I. focuses on. Even so called ‘neural networks’, created to replicate the brains synapse processing, simply process a set of inputs to produce some sort of output. The only thing holding us back from creating truly intelligent machines, A.I. proponents would argue, is our current processing power. “Intelligent machines” created today are bound by processing constraints, therefore their “intelligence” is constrained. However, without real understanding of the information being processed, can it be considered truly intelligent?

<

p>I would tend to agree with Jeff on his assertion that the A.I. community is going down the wrong path by equating behavior with intelligence. I think one of the reasons the A.I. community is going down this route of creating machines that behave intelligent is that our society itself often equates behavior with intelligence. Therefore, it would seem natural to replicate this assumption in the design of “intelligent” machines. A small example:

<

p>When I was in grade school ( grades 1–8 ), there was this boy, who we’ll call Sean, in the same class as me. He was considered very smart. I couldn’t figure out why people, including my teachers, treated us so different, even though we got similar grades. One day, I figured out why. It turns out that Sean would sit down and read the encyclopedia during recess everyday. He could talk about anything, as long as it was in the encyclopedia, and he read it recently. It turns out that people would use this as an example of why he was “smart” & “intelligent”. Simply by being able to read something and repeat it to other people was enough for most people to consider him intelligent.

<

p>In this case, it would be easy to “behave” intelligent, but is that really intelligence?

<

p>I think I was more intelligent that Sean because no only could I get the same grades as him, but I didn’t spend my recess in the corner.

Anywhoo…. The book is pretty good, though I am only just starting with the 3rd chapter. It will be interesting to see what direction Jeff takes A.I. and if any of it actually makes sense. Often times, people who are widely successful, like Jeff, think that their success bucking the system and “thinking differently” in one fields can be easily migrated to other fields.

Now playing: Various ArtistsNo Ordinary Love (Blacksky Mix) / Sherrie Lea from Thrive Mix 01 – Mike Ross