Stephen Baxter - The Science of Avatar

And then there are ethical and other doubts about the technique, as with so many other areas of modern medicine. For instance, babies can be “screened” in the womb for genetic conditions, possibly treated, or, perhaps, aborted if the parents choose. Many people will have doubts about where to draw the line in terms of such choices. Then there is the question of inheritance. There are two basic types of gene therapy. You can insert the therapeutic genes into the somatic cells of the patient—that is, the non-reproductive cells of the body. In this case any effects will be restricted to the patient only, and not passed on to any offspring. Or you can insert genes into germ cells —that is, reproductive cells, sperm or eggs. These changes would be heritable and can be passed on to future generations. These techniques are so controversial that in many countries, including the UK, tampering with the human germ line is a specific criminal offence.

One very unpleasant offshoot of gene therapy research could be “smart” biological weapons. You could target a specific group or individual with a particular DNA pattern, and trigger a natural or engineered disease. It must be hoped that this doesn’t occur to any SecOps think tanks on Pandora—but it is a possibility, since we know from the creation of the avatars that humans have to some extent mastered Na’vi genetics as well as their own.

The medical treatments discussed here are more or less at the experimental stage today. Perhaps the successful ones will be routinely available by the mid-twenty-second century. But it seems likely they will be costly. Aside from the evident cost of fixing Jake’s spinal injury, we see scientist Max Patel wearing glasses! If you can build an avatar, you’d think you could fix short-sightedness—but, obviously, only at the right price.

Another technological advance obvious in Hell’s Gate is computer technology.

Consider the Hell’s Gate Ops Centre control room. ( Avatar ’s creative team visited such locations as a real-world oil rig, the gigantic Noble Clyde Boudreaux in the Gulf of Mexico, to use as a model for interiors like this.) We see large-scale wraparound screens that respond to the touch and movement of the operator. In another instance, in the avatar lab, Max Patel swipes one tablet-like screen over another, taking an image to carry away with him to show Grace Augustine, as easily as he might pull a piece of paper from a pin-board. Three-dimensional displays are the norm, and there is an emphasis on graphic and tactile interactions, in an environment saturated with computing. These scenes recall recent experiments in “ubiquitous computing,” in which computers become embedded in the surroundings. Nokia’s Ubice is one prototype. In Microsoft’s Lightspace system, surfaces in a lecture room become screens for displaying documents and images; like Max you can pick up a virtual item from one display and move it to another.

The Ops Centre also features a holotable, with a continuously updated summary of conditions across RDA’s operations on Pandora. This is a very impressive, fully searchable holographic display, which Jake is able to reach into, tracing for Quaritch the internal structure of Hometree with his hands. Holography, the science of 3-D projection, is quite an old technology. The principles on which it is based were first set out in 1947 by the British physicist Dennis Gabor, who got a Nobel Prize for his trouble. Information about the amplitude and phase of light waves—that is, how intense they are and how they relate to each other—are stored as patterns of interference. Computer programs “ray-trace” back from these interference patterns to recreate the light rays that gave rise to those patterns, and so give the illusion that the object that emitted or reflected the light in the first place is present. Indeed, that “object” might only ever have existed in the electronic imagination of a computer.

Human-machine interaction (HMI) is the academic study of the interaction between people and computers. It is the intersection of a number of fields, from ergonomics and human factors to computer design. It arose partly because of bad examples of human-machine interfaces leading to calamity—for instance, it is thought that the Three Mile Island nuclear accident was partly due to operators struggling with a poor and confusing interface. HMI practitioners develop theories of interaction, come up with design methodologies and processes, and invent new kinds of interfaces and interaction techniques. A long-term goal is to minimise the barriers between a human’s cognitive model of what she wants to accomplish and the machine’s understanding of the task.

This makes sense in terms of what we see of the computer interfaces in Avatar , which seem a logical development from modern technology, our tablets and smart phones, with their applications which respond to touch, and can sense physical movements such as tipping and shaking thanks to internal accelerometers and GPS positional awareness. All of this builds an illusion that the computer applications are part of our physical world.

But if the human interfaces look familiar, current trends would suggest that we ought to anticipate huge advances in computer power by 2154.

“Moore’s law” is an empirical observation that thanks to technological advances and commercial pressure the speed of computer systems (as well as other parameters such as memory storage and relative cheapness) is growing exponentially. This was first described by Intel co-founder Gordon E. Moore, who in 1965 noted that the number of components in integrated circuits had doubled every year since the invention of such circuits in 1958. The doubling is cumulative, like compound interest, so in ten years the increase (two multiplied by itself ten times) would be over a thousandfold.

Similar studies based on other ways to calculate computing power give different values for the doubling time, but all of the same order of magnitude. Futurologist Ray Kurzweil has claimed the law has been working since the mechanical calculating machines of the early twentieth century. And it’s still working today, nearly half a century after Moore’s original paper. As of November 2010, according to the “TOP500” list that keeps a rank of such things, the most powerful non-distributed computer system in the world, a Chinese supercomputer called the Tianhe-1A (“the Milky Way”) was capable of around twenty-five hundred trillion elemental mathematical calculations per second (2.5 petaflops, in the jargon). The TOP500 list, maintained since 1993, confirms a version of Moore’s Law based on the big machines’ processing speeds, with a doubling time of fourteen months.

But Moore’s Law makes even mighty machines look dumb very quickly. With a fourteen-month doubling the Law should ensure that a laptop, presumably available for the same kind of comparative price as today, will pass the power of that big Chinese machine in a mere fifteen years . I won’t depress you here by telling you when the supercomputers, or indeed your phone , will become more powerful than your brain. We’ll consider that stuff in Chapter 32; it would certainly help with the tricky business of linking Jake to his avatar to have the whole process buffered by computers much more powerful than either brain.

Moore’s Law must have a limit beyond which it breaks down; in the end it will come up against fundamental physical limits. But by Avatar ’s mid-twenty-second century the world will surely be utterly saturated by extremely advanced computer technology. Just as today it’s in your TV and car and phone, by then we must anticipate that it will be everywhere, in your clothes, your home, in every gadget you use—even in the very fabric of your body, which might swarm with tiny smart medical-repair nano-robots.