Here's the exerpt that stood out to me:
Brooks: . . . This cognitive revolution is giving us a more accurate view of human nature that we are not only the rational, incentive based, the linguistic, logical parts of our mind, but we have other processes which are associational, which are emotional, and this is how we really navigate the world.
So to me this is just exciting in its own way, but also solves my problems, my problems of why we've had so many policy failures by giving us a more accurate view of human nature and how people are likely to respond to different situations.
CHARLIE ROSE: It's stunning.
And, you know, when I talk to all these people and frequently say "What's the one question you most want to answer?" It's always about the unconscious. There is this ultimate fascination with, as what you say, how much of what we do is unconscious, and how deeply influential it is in terms of all relationships we have.
DAVID BROOKS: Right. And to me, the metaphor that helps me understand it is the conscious is like a general, distanced from the world, looking down on it. The unconscious are like scouts, millions of them, permeating the world, going into other minds, going into an environment and sending back these emotional signals -- go, no, go.
And so when you look at a menu, you can make a choice of what you want, but you do not have a choice of whether you like broccoli. That's happening unconsciously. If you describe to me a story of incest, I have an immediate moral reaction to that. That's all done unconsciously.
And, you know, so one of my favorite little science experiments are people named Dennis are disproportionately likely to become dentists. People named Lawrence are disproportionately likely to become lawyers, because unconsciously we have a preference for things that are familiar and we follow those things.
And by now there are now 30,000, 40,000 neuroscientists in the world as well as other fields, and they're reshaping the way we understand how we navigate in this way, how these scouts send back signals. Sometime they send back very brilliant signals, sometimes they send back biased signals.
But it doesn't simplify life, but it gives you a sense of why we react or don't react the way we do.
CHARLIE ROSE: You've answered the question that I kept saying, why is David writing about this, one? And why is he at this cognitive conference and this other cognitive conference? Now I know, because of the book.
What can you tell me about the book? Can you tell me more? Is it about the brain? Is it about the unconscious?
DAVID BROOKS: I'm not a neuroscientist, so it's not really about the brain. So my goal is to go the whole book without using the word "amygdale," because there's all this stuff with MRIs and brain geography which is very cutting edge and it's not something I'm qualified to write about.
CHARLIE ROSE: But it is the frontier of science right now.
DAVID BROOKS: But I'm always -- for us outsiders, we really should be worried about it, because the brain has a hundred billion neurons and infinite, almost, numbers of connections, and sometimes I look at the brain scans, the nice color pictures, and you think, OK, they're flying over Los Angeles, they're looking at neighborhoods where the lights are on, and they're trying to guess at what people are talking about at the dinner table. It's phenomenally complicated . . .
Nonetheless, they don't solve philosophical problems. They don't give you a new philosophy of life. But they do confirm or validate some old philosophies.
If you thought that emotion was not separate from reason, that we were all fundamentally emotional creatures, then this confirms that, the importance of emotion. And so few if you felt we were fundamentally social creatures, then this confirms that, because we get dopamine surges when we have social conferences.
If you thought we were utilitarian, purely rational individualists, then this disconfirms all that. So it confirms certain -- it settles certain philosophical arguments, or at least biases you in one direction. And I found that just tremendously useful.
I thought his ideas about associations, the subconscious etc. are really pertinent to our class discussions over the last few weeks. I think that part of the reason why it is so hard for me to describe my artistic process or pinpoint what a thought is, is because it's subconscious. I'm not totally aware of what I do, but I feel like DARCI has to be aware of her every move. Maybe I'm wrong. I just wonder how close she can really mimic artificial intelligence because so much of our intelligence and functioning powers aren't conscious thoughts. Telling my heart to pump, my lungs to breathe, explaining why to put this color there and that color there . . . I can't explain how it happens, I just know it does.
That was a really interesting interview. It reminded me of a book I read called Blink, by Malcolm Gladwell. It’s basically about this idea that our subconscious is very active and learning using the stimulus around us all of the time. It’s called Blink because we often make very quick decisions in the blink of an eye, sometimes decisions that we barely even think about, and do so effectively because we are relying upon these subconscious processes to provide our conscious processes with accurate and useful information. Anyways, it’s really good. I recommend it to everyone. If you want to borrow it, I’ve got a copy.
ReplyDeleteMy personal crazy theory on the matter, is that these subconscious processes are not terribly unlike machine learning algorithms (like those implemented in DARCI). This would mean that our subconscious is learning and making decisions in the same sense that we say DARCI is making decisions. We are inclined to only call something a choice if it reaches our consciousness—but what about all of these unconscious decisions. Do they count? It seems like the most complex decisions, the ones that need to amalgamate a lot of information, are reserved for conscious reasoning. So these seem like the most interesting choices to me.
Speaking theologically, maybe the conscious part of our thinking is when the brain brings the conclusion of many processes to our Intelligence’s attention and says something like this: “OK, I’ve analyzed a whole bunch of data. I’ve found these associations and made these predictions. There’s a whole lot of information that I’m not bringing to your attention because it would be too much to think about at once. But here is the best info, now make a decision!”
So is it possible to simulate this by building a super process (a General) that manages a bunch of subservient processes?
I think it is possible. What you describe seems to be similar to concept of meta-learning. In fact, I don't know why I haven't seen more of that type of research in machine learning. Combining several different learning tasks to solve a higher task seems like that's what we want/need to do in the end. Come to think of it, DARCI actually does a little bit of that when she learns to describe images with adjectives, then uses that knowledge to solve the higher task of creating new images.
ReplyDelete