Robot recognition

It is unsettlingly significant or pathetically kitsch that the first robot ‘scientist’ is called Adam. In either case, Adam has been accredited with discovering the gene in the DNA sequence of yeast bacterium related to its growth and more importantly, beating his human counterparts to the discovery. While this can largely be attributed to the repetitive nature of the task, it is nevertheless a triumph of robotics and a milestone on the path to a genuine artificial intelligence.

AI has always been an edgy science. The West’s hang-up with Judeo-Christian beliefs has been blamed for the apprehension that such technology seeks to play god. Meanwhile, Hollywood can also be held accountable for introducing its own dose of blind cynicism into the debate (films such as The Terminator tend to treat the advent of AI with fear and uncertainty). Yet the Japanese have been far more open minded in embracing the possibility of a robotic future, driven in equal measure by their advanced technological society and Shinto belief system.

Though the prospect of adding subjectivity to an artificial creation’s thinking process remains distant, science is taking significant steps. The key to this reality however lies in learning more about ourselves. Research is currently focusing on creating cybernetic replacements for damaged parts of the human body; from bridging damaged parts of the brain through computer chips to replacing amputated limbs. Results so far have been basic, but none the less successful.

One major problem has been creating technology small enough to match that of the human body, particularly in terms of power supplies. Yet scientists such as Dr Ted Berger of the University of Southern California have already created computer chips that can replicate the role of the hippocampus (the part of the brain responsible for creating short-term memory), while scientists at MIT are at the latter stages of developing a battery the size of a human cell.

While various human body and brain functions have been replicated, the idea of getting them all to work together – especially without an organic base to build upon (a potential minefield for ethics) – remains a major challenge. Still not enough is known about the human brain to be able to replicate its vast functionality, with even our best attempts staggeringly primitive.

Yet this is AI in a conventional sense. Intelligence could develop in other ways. Consider the basic internet search engines. The fact that the same search terms sometime yield different results suggests priority, or choice, by the search algorithm. Even if caused by computer servers acting at slightly different times, the result represents a basic example of one of the fundamental faculties of intelligent life.

Consider the internet in general: a huge, interconnected series of signals passing from one processor to another to convey a piece of sensory information. Is this not also how the human brain works to provide us with the information we need?

Both of these interpretations come from an anthropomorphic view of intelligence. Yet there could be new forms of intelligence. Indeed, signs of basic artificial intelligence could be in existing technology and we might not even notice them, much as we may not recognise signals from space being from other forms of intelligent life.

Putting aside this mindboggling concept however, it’s almost certain that narcissism will deliver AI. Adam is the first of many robots to come who will eventually surpass humans in their achievement (One might also mention Watson, who recently won Jeopardy).

With Moore’s law having proved startlingly accurate thus far, independent AI is certainly likely to occur sooner than we might think.