rtificial intelligence—the ability of machines to perform complex tasks through programming that enables decision making—is already part of our everyday life. It plays an increasingly important role in industrial production processes and in banking, and some equity funds are already successfully managed by artificial intelligence.
A distinction is often made between “weak AI,” as in problem-solving machines and programs limited to special tasks such as playing chess and other games, and “strong AI,” which involves some degree of self-awareness or even consciousness. It is anticipated that around the year 2015, the computing power of personal computers will have reached the computing power of our brains. The question is whether it will be possible to program artificial consciousness.
As we do not have a scientific theory of consciousness, and most scientists and philosophers think that such a theory will never be available, it is not a straightforward task to supply a computer with consciousness. The philosopher John Searle has argued that computer-simulated consciousness will never be true consciousness, and the physicist Roger Penrose has proposed that human consciousness involves processes on the quantum level of physics that are beyond our present knowledge. Most neuroscientists, however, see no way for the information processing function of neurons (nerve cells) to access single quantum mechanical processes, and others, such as Ray Kurzweil, a pioneer in fields including optical character recognition and text-to-speech synthesis, believe that we will be able to upload the human mind to a computer within a few decades.
The most promising approach to reaching a human level of intelligence and even selfawareness artificially is to copy the way the brain works. The information processing function of neurons has been studied extensively, and there are now software models available that simulate the behavior of these cells to a very high degree of accuracy. If one combines a large number of these artificial neurons into neural nets, one gets functional units that work very similarly to units of the human brain. These units can be trained to perform specific tasks, such as the recognition of speech or visual patterns. It is not necessary to copy the hardware of the human brain exactly, just to program the function of a neural net in a standard computer. For the optimization of the connections between the virtual neurons, evolutionary algorithms have been found to be very useful.
The first entertainment robots (such as Sony’s “AIBO” and Wowee’s “Robosapien”) are now produced in large numbers. Although they can perform only very simple tasks at the moment, we will see continuous improvement to their intelligence. As they acquire the ability to perform more and more complex tasks, their application will extend beyond entertainment to such tasks as assistance in housekeeping and health care.
The shrinking size and reduced power consumption of processor chips are also yielding new possibilities for neural implants. Already most deaf patients can be treated with cochlear implants—not simply improved hearing aids that amplify sounds, but complex devices that perform a fast Fourier transform calculation of the received sound within milliseconds and distribute the resulting spectral function directly to the cochlea by way of tiny electrodes. Initial experiments have also been done with retinal chip implants for the blind, and another anticipated application is the development of neural implants for paraplegic patients.
Within a few decades we will be able not only to “repair” humans with a variety of neural implants and artificial organs, we will be able to increase human capabilities beyond their biological limits, resulting in the so-called Cyborg, or cybernetic organism. In addition,
if artificial consciousness proves possible, we will be able to construct conscious robots. A civilization that includes such new kinds of beings is referred to as transhuman, and should a transhuman civilization emerge, humankind may find itself in the novel position of no longer being the summit of creation. Designers of transhuman beings will have tremendous power, the source of serious ethical concerns. Critics of such a development often cite Aldous Huxley’s Brave New World, a dystopian view of a future dominated by state-controlled technology. What such critics overlook is that the main problem in Huxley’s world is not the technology but the totalitarian state.