I have to own up that the first time I saw the cinema ad for 'Godfather: The Game', I mistook it (just for a second) for actual footage of Brando and was puzzled because I couldn't remember where in the film it came from. OK, so the actual gameplay is by all accounts crap, and OK its startling realism relies mostly on a single trick - a deep and all-pervading chiaroscuro that hides all the rough edges and angles (I'd guess the designers hung around the Caravaggio exhibition a lot). Nevertheless it's a testament to how close the computer-generated virtual world can now get to the virtual world of the Hollywood movie, and I'm sure the gap will continue to close at ever faster pace.
Notice that I didn't use the phrase 'real world' at all in the above sentence, because of course the world of live-action movies is every bit as virtual as any computer game. They're both just representations, neither of which can convincingly simulate a third dimension (those helmets never did catch on) nor solidity. Most tellingly both operate only on two of our senses, sight and hearing, which is no coincidence since those two senses are rather different from the rest (I almost wrote 'the other three' but a horrid vision of a leering Stephen Fry and that awful QI klaxon noise stopped me just in time). A naive critic of virtual reality might complain that computers can only show us the surface of things, but of course so do our own senses. They deliver to us images of an external world which our brains then interpret as composed of remote objects surrounding us and distributed in space.
In the case of both sight and hearing, no contact with these external objects is necessary: only reflected or emitted light rays pass through our corneas, and only aerial compression waves batter our eardrums, never the things that emit them (hopefully). By contrast our other senses require material contact with the external object, some part of which may actually penetrate our bodily perimeter: tasting and smelling both involve ingesting (small) quantities of foreign substances, and the difference between a touch and a skin-piercing wound is only a few pounds of extra pressure. The world these senses depict (the world of the deaf-and-blind) lies at much closer range, it lacks vistas or landscapes. However it doesn't for that reason lack in powerful emotional stimuli: on the contrary these senses carry more emotional charge than either sight or sound as they depend on us having a body. Sight and sound are fully detached from the objects they detect, so a being without a body like a computer can far more easily simulate them.
Computer graphics will soon overtake the performance of the human eye (sound processing has already overtaken the performance of the ear) to create images that we can't distinguish from reality. Our eyes and ears are after all digital devices too - the retinal rods and cones, and the ear's cochlear hairs are their equivalent of CCDs (though they're not binary as the cyborg crowd appear to believe). Their bandwidth is fairly modest and well within the abilities of modern silicon. However our brain does much, much more than just process pictures and noises: it integrates all our sensory inputs to make us believe in a persistent, solid world that contains other agents besides ourselves who may have intentions toward us. We even have special brain circuits for interpreting other animal's facial expressions to intuit whether they're angry, sad, or perhaps fancy us. Nothing we see is emotionally neutral, it's either good for us or bad for us, and gets tagged as such in memory. And not everything we see necessarily exists: the brain continually calls on memory to fill in those parts of the world it lacks the bandwidth to monitor in real time, to say nothing of dreams, daydreams, imagination and hallucinations. Just like a virtual reality computer, the brain can dip into its database to create worlds from the inside, in addition to perceiving the one outside.
Now even if it were possible, it wouldn't be a good idea to make computers emulate all of our emotion-processing abilities. Messrs Kubrick and Clarke knew this decades ago when they created 2001's HAL. You'd have to give the machine a notion of its own good, which would then distract it from doing our bidding. (The advantage of dumb computers is that you don't need to feel guilty about abusing them on those occasions when they don't do our bidding). Nowadays some 'affective-computing' researchers are experimenting with friendly user interfaces that use sampled human facial expressions and exploit the brain's own expression interpreter rather than trying to mimic it via AI. Microsoft's infuriating Clippy the paperclip would seem to have taught them nothing. Face it, it's hard enough to cast real actors who don't irritate, so what chance virtual ones?
[PULL OUT
The average human being's senses gather data at approximately 11Mbits/sec, divided between the individual senses as follows:
Sight 10Mbit/sec
Touch 1Mbit/sec
Hearing 100Kbit/sec
Smell 100Kbit/sec
Taste 1Kbit/sec
Cognitive science experiments suggest that the conscious mind is only capable of processing about 50bits/sec, so most processing must proceed unconsciously, to achieve a stunning degree of data compression.]
Dick Pountain/11 April 2006/15:26/Idealog 141
SIGHTS AND SOUNDS
I have to own up that the first time I saw the cinema ad for 'Godfather: The Game', I mistook it (just for a second) for actual footage of Brando and was puzzled because I couldn't remember where in the film it came from. OK, so the actual gameplay is by all accounts crap, and OK its startling realism relies mostly on a single trick - a deep and all-pervading chiaroscuro that hides all the rough edges and angles (I'd guess the designers hung around the Caravaggio exhibition a lot). Nevertheless it's a testament to how close the computer-generated virtual world can now get to the virtual world of the Hollywood movie, and I'm sure the gap will continue to close at ever faster pace.
Notice that I didn't use the phrase 'real world' at all in the above sentence, because of course the world of live-action movies is every bit as virtual as any computer game. They're both just representations, neither of which can convincingly simulate a third dimension (those helmets never did catch on) nor solidity. Most tellingly both operate only on two of our senses, sight and hearing, which is no coincidence since those two senses are rather different from the rest (I almost wrote 'the other three' but a horrid vision of a leering Stephen Fry and that awful QI klaxon noise stopped me just in time). A naive critic of virtual reality might complain that computers can only show us the surface of things, but of course so do our own senses. They deliver to us images of an external world which our brains then interpret as composed of remote objects surrounding us and distributed in space.
In the case of both sight and hearing, no contact with these external objects is necessary: only reflected or emitted light rays pass through our corneas, and only aerial compression waves batter our eardrums, never the things that emit them (hopefully). By contrast our other senses require material contact with the external object, some part of which may actually penetrate our bodily perimeter: tasting and smelling both involve ingesting (small) quantities of foreign substances, and the difference between a touch and a skin-piercing wound is only a few pounds of extra pressure. The world these senses depict (the world of the deaf-and-blind) lies at much closer range, it lacks vistas or landscapes. However it doesn't for that reason lack in powerful emotional stimuli: on the contrary these senses carry more emotional charge than either sight or sound as they depend on us having a body. Sight and sound are fully detached from the objects they detect, so a being without a body like a computer can far more easily simulate them.
Computer graphics will soon overtake the performance of the human eye (sound processing has already overtaken the performance of the ear) to create images that we can't distinguish from reality. Our eyes and ears are after all digital devices too - the retinal rods and cones, and the ear's cochlear hairs are their equivalent of CCDs (though they're not binary as the cyborg crowd appear to believe). Their bandwidth is fairly modest and well within the abilities of modern silicon. However our brain does much, much more than just process pictures and noises: it integrates all our sensory inputs to make us believe in a persistent, solid world that contains other agents besides ourselves who may have intentions toward us. We even have special brain circuits for interpreting other animal's facial expressions to intuit whether they're angry, sad, or perhaps fancy us. Nothing we see is emotionally neutral, it's either good for us or bad for us, and gets tagged as such in memory. And not everything we see necessarily exists: the brain continually calls on memory to fill in those parts of the world it lacks the bandwidth to monitor in real time, to say nothing of dreams, daydreams, imagination and hallucinations. Just like a virtual reality computer, the brain can dip into its database to create worlds from the inside, in addition to perceiving the one outside.
Now even if it were possible, it wouldn't be a good idea to make computers emulate all of our emotion-processing abilities. Messrs Kubrick and Clarke knew this decades ago when they created 2001's HAL. You'd have to give the machine a notion of its own good, which would then distract it from doing our bidding. (The advantage of dumb computers is that you don't need to feel guilty about abusing them on those occasions when they don't do our bidding). Nowadays some 'affective-computing' researchers are experimenting with friendly user interfaces that use sampled human facial expressions and exploit the brain's own expression interpreter rather than trying to mimic it via AI. Microsoft's infuriating Clippy the paperclip would seem to have taught them nothing. Face it, it's hard enough to cast real actors who don't irritate, so what chance virtual ones?
[PULL OUT
The average human being's senses gather data at approximately 11Mbits/sec, divided between the individual senses as follows:
Sight 10Mbit/sec
Touch 1Mbit/sec
Hearing 100Kbit/sec
Smell 100Kbit/sec
Taste 1Kbit/sec
Cognitive science experiments suggest that the conscious mind is only capable of processing about 50bits/sec, so most processing must proceed unconsciously, to achieve a stunning degree of data compression.]
No comments:
Post a Comment