Warning: session_start(): open(/var/lib/php/session/sess_am3ff01prqj6bakrpnflu7ms4r, O_RDWR) failed: No space left on device (28) in /var/www/html/includes/config.php on line 34

Warning: session_start(): Failed to read session data: files (path: /var/lib/php/session) in /var/www/html/includes/config.php on line 34
addabook - On Intelligence
addabook home timeline gallery
signup or login
On Intelligence
How a New Understanding of the Brain Will Lead to the Creation of Truly Intelligent Machines
Jeff Hawkins
read on February 1, 2013

Getting right to it, Jeff Hawkins seems like a snob - it was hard to get over that. I spent most of the first half of the book looking for reasons to disagree with him, just because he seems like the kind of guy you'd really hate to get stuck in an elevator with. I'll bet him and Nasim Taleb would get along just great. Anyway, despite my best (and his own) efforts, Hawkins doesn't offer too much to disagree with. I expected this book to actually disagree with Kurzweil's take on electronic intelligence, and I think Hawkins would say that it does - but frankly, I don't. I think they're both largely saying the same thing, but Hawkins is trying to make distinctions which really don't matter that much. For instance, he spends a terribly long amount of time talking about how physical computer hardware isn't like the brain, how the analogy that folks like Kurzweil is wrong. Yeah, duh. Who cares? That doesn't say anything about our ability to recreate intelligence in software systems. Hawkins is just trying to stir the pot. Whatever. There are four things I specifically wanted to call out:

  • First, Hawkins really digs into the "Chinese Room" metaphor/thought experiment. This, roughly, tries to figure out if machine translation is intelligent. If you had a Chinese guy in one room, and an English-only speaker in the other, and the Chinese guy communicated with the English guy only by passing him (through a slot in the wall) papers with Chinese writings on them - and the English guy had no idea what they meant, but was able to look them up in a huge manual that told him exactly what to write back on new sheets of paper, such that when the Chinese guy read the Chinese writing that the English guy and his manuals had created the Chinese guy was sufficiently convinced that the person on the other side spoke Chinese - would you then consider the English guy to "understand" Chinese? If not the guy, did the manuals understand Chinese? Did the whole system (the man and the room, etc)? During which part of the process was their intelligence? What is the nature of intelligence? Et cetera. These were interesting things to think about - which I'm actually not done thinking about.
  • Back in Moonwalking With Einstein Joshua Foer wrote that how much you remember of a new experience is a function of how much you already know. This is because you remember things in context, so you need to have a contextual framework to stick things onto in your head. This made perfect intuitive sense to me when I read it, but it also was something I had never thought of or realized and I thought it was profoundly important, and it very much changed my reading habits and motivated me to start researching a much broader scope of general knowledge, so as to increase my general retention of information throughout life. Hawkins argues the exact opposite. Hawkins builds out a framework for the brain wherein only novel events actually make it all the way through a complex hierarchy and into your hippocampus to be stored as long term memory. In fact, he argues (anecdotally, and seemingly in jest, but not entirely) that part of the reason you remember less as you age is because so many fewer things are novel. Your brain already knows most of what you experience, so it doesn't make it up the pyramid to the hippocampus. I have no idea if there is merit to this, but it was interesting to see two theories about that same thing that are so diametrically opposed, yet both make such good intuitive sense. I'm certainly rooting for Foer.
  • Hawkins brings up synesthesia, an incredible sounding brain disorder where your senses overlap. Folks with this condition may think that a certain taste is red, or that a certain word is rough and brittle. They just cross their senses in ways that don't make any sense at all to normal people, but make perfect sense to them. I've read about this before in other books but was happy to see it pop up again here and wanted to record it. I'm sure this can be a difficult and possibly debilitating disorder to live with, but some part of me is happy it exists. I like the idea that there are people out there who experience classical music as green, and think that rock music tastes like mustard. I can't explain why, but I with all due respect I feel like overall we're better for it, to have those folks around.
  • Hawkins proposes that the core nature of intelligence is the ability to predict the future. He proposes this like it's his own theorem, like he owns the rights to this idea. Towards the beginning of the book, he asks the reader what they think the nature of intelligence is - and without skipping a beat I thought to myself "ability to predict the future". Now look, I'm fairly certain that that is something I've come across in other neuroscience books before, I'm not trying to take credit - but I haven't read any of his books before. I just thought it was a jackass thing to try and claim that idea for himself.

Author Bio:

Jeffrey Hawkins (/ˈhɔːkɪnz/; born June 1, 1957) is the American founder of Palm Computing (where he invented the PalmPilot) and Handspring (where he invented the Treo). He has since turned to work on neuroscience full-time, founded the Redwood Center for Theoretical Neuroscience (formerly the Redwood Neuroscience Institute) in 2002, founded Numenta in 2005 and published On Intelligence describing his memory-prediction framework theory of the brain. In 2003 he was elected as a member of the National Academy of Engineering "for the creation of the hand-held computing paradigm and the creation of the first commercially successful example of a hand-held computing device." Hawkins also serves on the Advisory Board of the Secular Coalition for America and offers advice to the coalition on the acceptance and inclusion of nontheism in American life.