Sensual Experience



Modern day gamemakers are constantly working on higher-resolution, more realistic graphics; more immersive sounds—but they're a long long way off real life.

Pressing the "W" key to run forward as a graphic of a hungry tiger bounds behind you, just doesn't seem quite as sensual as running frantically across the savanna with your own legs, breathing in huge gasps and pumping your arms as the sun beats down on your shoulders, the grass brushes your shins, and the air whips around you with the wind of your passage.

Don't mistake me for a luddite; I'm not saying the technology can't get that good.  I'm saying it hasn't gotten that good yet.

Failing to escape the computer tiger would also have fewer long-term consequences than failing to escape a biological tiger—it would be less a part of the total story of your life—meaning you're also likely to be less emotionally involved.  But that's a topic for another post.  Today's post is just about the sensual quality of the experience.

Sensual experience isn't a question of some mysterious quality that only the "real world" possesses.  A computer screen is as real as a tiger, after all.  Whatever is, is real.

But the pattern of the pseudo-tiger, inside the computer chip, is nowhere near as complex as a biological tiger; it offers far fewer modes in which to interact.  And the sensory bandwidth between you and the computer's pseudo-world is relatively low; and the information passing along it isn't in quite the right format.

It's not a question of computer tigers being "virtual" or "simulated", and therefore somehow a separate magisterium. But with present technology, and the way your brain is presently set up, you'd have a lot more neurons involved in running away from a biological tiger.

Running would fill your whole vision with motion, not just a flat rectangular screen—which translates into more square centimeters of visual cortex getting actively engaged.

The graphics on a computer monitor try to trigger your sense of spatial motion (residing in the parietal cortex, btw).  But they're presenting the information differently from its native format —without binocular vision, for example, and without your vestibular senses indicating true motion.  So the sense of motion isn't likely to be quite the same, what it would be if you were running.

And there's the sense of touch that indicates the wind on your skin; and the proprioceptive sensors that respond to the position of your limbs; and the nerves that record the strain on your muscles.  There's a whole strip of sensorimotor cortex running along the top of your brain, that would be much more intensely involved in "real" running.

It's a very old observation, that Homo sapiens was made to hunt and gather on the savanna, rather than work in an office.  Civilization and its discontents...  But alienation needs a causal mechanism; it doesn't just happen by magic.  Physics is physics, so it's not that one environment is less real than another.  But our brains are more adapted to interfacing with jungles than computer code.

Writing a complicated computer program carries its own triumphs and failures, heights of exultation and pits of despair.  But is it the same sort of sensual experience as, say, riding a motorcycle?  I've never actually ridden a motorcycle, but I expect not.

I've experienced the exhilaration of getting a program right on the dozenth try after finally spotting the problem.  I doubt a random moment of a motorcycle ride actually feels better than that.  But still, my hunter-gatherer ancestors never wrote computer programs.  And so my mind's grasp on code is maintained using more rarefied, more abstract, more general capabilities—which means less sensual involvement.

Doesn't computer programming deserve to be as much of a sensual experience as motorcycle riding?  Some time ago, a relative once asked me if I thought that computer programming could use all my talents; I at once replied, "There is no limit to the talent you can use in computer programming."  It's as close as human beings have ever come to playing with the raw stuff of creation—but our grasp on it is too distant from the jungle.  All our involvement is through letters on a computer screen.  I win, and I'm happy, but there's no wind on my face.

If only my ancestors back to the level of my last common ancestor with a mouse, had constantly faced the challenge of writing computer programs!  Then I would have brain areas suited to the task, and programming computers would be more of a sensual experience...

Perhaps it's not too late to fix the mistake?

If there were something around that was smart enough to rewrite human brains without breaking them—not a trivial amount of smartness—then it would be possible to expand the range of things that are sensually fun.

Not just novel challenges, but novel high-bandwidth senses and corresponding new brain areas.  Widening the sensorium to include new vivid, detailed experiences.  And not neglecting the other half of the equation, high-bandwidth motor connections—new motor brain areas, to control with subtlety our new limbs (the parts of the process that we control as our direct handles on it).

There's a story—old now, but I remember how exciting it was when the news first came out—about a brain-computer interface for a "locked-in" patient (who could previously only move his eyes), connecting control of a computer cursor directly to neurons in his visual cortex.  It took some training at first for him to use the cursor—he started out by trying to move his paralyzed arm, which was the part of the motor cortex they were interfacing, and watched as the cursor jerked around on the screen.  But after a while, they asked the patient, "What does it feel like?" and the patient replied, "It doesn't feel like anything."  He just controlled the cursor the same sort of way he would have controlled a finger, except that it wasn't a finger, it was a cursor.

Like most brain modifications, adding new senses is not something to be done lightly.  Sensual experience too easily renders a task involving.

Consider taste buds.  Recognizing the taste of the same food on different occasions was very important to our ancestors—it was how they learned what to eat, that extracted regularity.  And our ancestors also got helpful reinforcement from their taste buds about what to eat—reinforcement which is now worse than useless, because of the marketing incentive to reverse-engineer tastiness using artificial substances.  By now, it's probably true that at least some people have eaten 162,329 potato chips in their lifetimes.  That's even less novelty and challenge than carving 162,329 table legs.

I'm not saying we should try to eliminate our senses of taste.  There's a lot to be said for grandfathering in the senses we started with—it preserves our existing life memories, for example.  Once you realize how easy it would be for a mind to collapse into a pleasure center, you start to respect the "complications" of your goal system a lot more, and be more wary around "simplifications".

But I do want to nudge people into adopting something of a questioning attitude toward the senses we have now, rather than assuming that the existing senses are The Way Things Have Been And Will Always Be.  A sex organ bears thousands of densely packed nerves for signal strength, but that signal—however strong—isn't as complicated as the sensations sent out by taste buds.  Is that really appropriate for one of the most interesting parts of human existence?  That even a novice chef can create a wider variety of taste sensations for your tongue, than—well, I'd better stop there.  But from a fun-theoretic standpoint, the existing setup is wildly unbalanced in a lot of ways.  It wasn't designed for the sake of eudaimonia.

I conclude with the following cautionary quote from an old IRC conversation, as a reminder that maybe not everything should be a sensual experience:

<MRAmes> I want a sensory modality for regular expressions.


Part of The Fun Theory Sequence

Next post: "Living By Your Own Strength"

Previous post: "Continuous Improvement"