Выбрать главу

To make such changes at earlier stages would involve too many details. If you only recorded a picture-like image, then it would be hard to change any part. But at the higher ‘semantic’ levels, you can more easily make more meaningful ways changes. For example, when you describe “a lying-down block supported by two upright blocks,” you need not specify the viewer’s perspective, or even say which parts of the scene are in view. Consequently that same description applies equally well to all these views:

If we substitute ‘object’ for the word ‘block,’ then our network would describe yet more situations, including these:

This shows how convenient are ‘abstract’ descriptions. Sometimes the word ‘abstract’ is used to mean ‘intellectually difficult’—but here it has almost the opposite sense: abstract descriptions are simpler when they suppress details that are not relevant. Of course, descriptions must not be too abstract: as when you ask someone for advice, and they give you a useless reply like, “If you want something, do what will get it for you.”

We’ve discussed how we might imagine visual scenes by constructing “simuli” inside our minds. We do similar things in other realms. Perhaps some chefs imagine new textures and tastes by changing their lower-level sensory states—and perhaps some composers imagine the sounds of new kinds of instrumentations—but such thinkers might also achieve such effects by making smaller changes at higher levels of representation, and thus evoke delight or disgust without constructing low-level details of those imagined musics or meals.

Drama Critic: I can clearly recollect how I felt after attending a certain performance, but I can’t remember any details at all of what that dreadful play was about.

To discuss this, we’ll coin a new word by combining “simulate” and “stimulus.” A simulus is a counterfeit perception caused by changing a mental representation. Thus in the Challenger scene of §4-7, we saw how a simulus of defeat could be used to evoke a feeling of Anger. To do this, it might suffice to represent no more than a sneer on one’s enemy’s face—with no other features of that face—for one can get by with the simplest kinds of descriptions by using the highest level abstractions.

Visualizer: When I think about my cat, its image is filled with so many details that I can visualize every hair. Would there not be a major advantage to making a real, pictorial image.[89]

Perhaps when you first imagine that cat, its surface has only a ‘furry texture’—and only when you ‘zoom in’ on it do you add more details to your mental representation. However, this could happen so quickly that you have no sense of it happening, and then it may seem to you as though you saw all those details at once. This could be an example of the illusion we mentioned in chapter §4:

The Immanence Illusion: When your questions get answered before you have asked them, it will seem that you’re already aware of those answers.

The Immanence Illusion applies not only to scenes that we imagine; we never see real scenes ‘all at once’, either, because we don’t perceive most fine details until some parts of our minds make requests for them. Indeed, recent experiments suggest that our inner descriptions of visual scenes are rarely updated in real time.[90] Chapters §6 and §8 will describe a scheme called “Panalogy” which might help to explain how our brains get such answers so rapidly.

∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞

§5-9. Prediction Machines

William James: Try to feel as if you were crooking your finger, whilst keeping it straight. In a minute it will fairly tingle with the imaginary change of position; yet it will not sensibly move, because ‘it is not really moving’ is also a part of what you have in mind. Drop this idea, think of the movement purely and simply, with all brakes off; and, presto! It takes place with no effort at all.

[The Principles of Psychology, 1890, p527.]

Everyone can think about things, without performing actions—as when Carol imagined moving those blocks. But how did she manage to do that? You, yourself could now close your eyes, lean back in your chair, and indulge in your own dreams and fantasies, reflect upon your motives and goals, or try to predict what will happen next.

Now, here is how we could make a machine that does that same sort of thing, by predicting the outcomes of various actions. Let’s assume that it has some rules like these.

Then we’ll give our machine—let’s call it Seer—a way to replace what it currently sees by the prediction described by this rule. Then when Seer is in situation A, and then considers doing action X, this will cause Seer then to ‘imagine’ that it is now in a situation like B.

I included that pair of “Suppressor Bands” for two separate reasons. First, when Seer imagines that future condition B, we do not want this to be quickly replaced by a description of the actual, present condition A. Second, we do not yet want Seer to perform action X, because it might want to consider some other options before it makes a final decision. So, Seer can use those suppressor bands to detach itself from the outside world—and this enables it to “stop and think” before it decides which action to take.[91]

By repeating this kind of operation, Seer could use such prediction-chains to simulate what happens in ‘virtual worlds.’ Of course, for Seer to be able to make such predictions it must be able to use the kinds of search we described in §5-3 to simulate (and then compare) the effects of difference courses of action before deciding which one to adopt. This will need additional memory, as well as other kinds of machinery. Still, anyone who has played a modern computer game can see how advanced has become the art of building virtual worlds inside machines.

I expect that in the next few years, we’ll discover structures like those in this diagram in various parts of human brains. How did our brains evolve these abilities? The species of primates that preceded us must have had some structures like these, which they could think several steps ahead. But then, a few million years ago, that system appears to have rapidly grown, as the frontal lobes of our brains developed their present great size and complexity—and this must have been a crucial step toward the growth of our human intelligence.

Summary

This chapter described some structures and processes that might do some of the things that people do. We outlined a sequence of levels at which we can use increasingly ways to think

However, we have suggested rather few details about what happens at each of those levels. Later I will suggest that our systems mainly work, at each of those various cognitive levels, by constantly reacting to the particular kind of troubles they meet—by switching to more appropriate Ways to Think. We’ll represent this Model of Mind by using this simple diagram:

вернуться

89

Some persons claim to imagine scenes as though looking at a photograph, whereas other persons report no such vivid experiences. However, some studies appear to show that both are equally good at recalling details of remembered scenes.

вернуться

90

See, for example, http://www.usd.edu/psyc301/Rensink.htm and http://nivea.psycho.univ-paris5.fr/Mudsplash/Nature_Supp_Inf/Movies/Movie_List.html.

вернуться

91

This prediction scheme appears in section §6-7 of my 1953 PhD thesis, “Neural-Analog Networks and the Brain-Model Problem, Mathematics Dept., Princeton University, Dec. 1953. At that time, I had heard that there were ‘suppressor bands’ like the one in my diagram, at the margins of some cortical areas. These seem to have vanished from more recent texts; perhaps some brain researchers could find them again.