Artificial Intelligence: from science fiction to reality (appeared on L'Express, Mauritius, May 2011)



As I furiously channel-surfed recently, in search of something palatable, I froze on the unmistakable and soothing voice of HAL in 2001: A Space Odyssey. 2001 is one of those timeless movie classics whose re-runs always bring something new and thought-provoking. Despite mixed reviews, many acknowledge 2001 as one of the greatest movies ever made. Directed by Stanley Kubrick (who co-authored the story with science fiction master, Arthur Clarke), 2001 dabbles with artificial intelligence (AI) and extraterrestrial life. 

What distinguishes it from other movies of this genre is how well the apparent surrealism is grounded into good science. The main character in 2001 is a ubiquitous computer on a spacecraft. HAL could see, listen, talk and understand his human “colleagues” anywhere on board. In fact, HAL was the “man-in-charge”, leading a secret mission to search for life in space. Another famous omnipresent, bad-boy computer is The Matrix, where self-aware, sentient machines create a simulated reality to pacify the human population, while using the energy in human bodies to survive.

Stanley Kubrick later wrote another movie called AI, completed by Steven Spielberg after Kubrick’s demise in 1999. This time, the computers were more anthropomorphic, such as Jude Law as a gigolo-robot. Or Haley Osment, as a heart-breaking child-robot with love for his mother that lasted millennia - talking of eternal love, as long as there is power supply.

There is no dearth of fictional characters to portray good and evil robots, androids and gynoids (woman robots). From Kalevala, the prehistoric woman forged out of gold in Finnish myth, to Governator Arnold Schwarzenegger in the Terminator series – or my favourites, C-3PO and R2-D2, the odd couple of Star Wars, and affectionate names for my twin kids!

HAL was a good excuse to reassess the progress of artificial intelligence. Is AI for real, or just a fad in academic and science fiction circles? There was a lot of excitement about the promises of AI when the field started in 1950s. But this was followed by an “AI winter”, when many were disappointed and research funding ran dry. Sceptics argue that simulating humans and their intelligence is decidedly difficult or even impossible – given technological as well as existential barriers. 

We are still a long way from self-aware humanoids serving us coffee and doing our taxes. Nonetheless, there are some remarkable feats of AI, often taken for granted. Like Deep Blue’s historic victory against chess world champion Gary Kasparov in 1997. Or earlier this year, two champions at Jeopardy (the “Questions pour un champion” of USA) were defeated by another IBM computer dubbed Watson. And  there are countless practical applications where computers have accomplished tasks in human (or super-human) fashion, well beyond complicated arithmetic: taking complex decisions on assembly lines or cleaning your house (robotics), diagnosing issues and suggesting solutions in knowledge-intensive fields such as medicine or law (expert systems), seeing the world and interpreting images (computer vision), anticipating the effects of moves before taking them, even on remote planets (planning), programs that search online and extract  knowledge (intelligent agents and data mining),  or even artificial pets and robots to comfort you (emotional agents and affective computing). 

AI seems to be reviving gradually. Some even foresee that AI will soon alter history. Scientists like Raymond Kurzweil and Vernor Vinge talk of a “singularity”, a turning point when machine intelligence would surpass human intelligence and take over the process of technological invention, with unpredictable consequences. Will reality be eventually stranger than fiction? Who knows? Maybe some computers will. (To be continued).

No comments:

Post a Comment