AI Then and Now
Like many scientific endeavors, the beginnings of artificial intelligence are can be traced back to visions of an imagined future, dreams of what may one day be possible. Science fiction writing has provided us with many ideas which have been realized and others that inspire research. 1950 saw both the publication Isaac Asimov' influential I, Robot, which contained the three laws of robotics (or artificial intelligence, and Allan Turing's paper Computing Machinery and Intelligence, which proposed the Turing Test, a means to assess whether a machine could emulate human responses enough to pass for a human. Remember, these came out just as the first computers were invented, machines that comprised thousands of vacuum tubes, occupied the space of a small warehouse, had no audio or graphical display, and the processing capability of a good scientific calculator today.
Humans become intelligent because we learn and this is key to the development of AI. As early as the the 1950's, a computer was getting better at playing checkers from playing against humans and computer opponents.
Checker's big brother, chess has played a major role in AI development, with various machines and computers programmed to rival chess champions. A milestone was reached in 1997 when IBM's Deep Blue program defeated reigning world champion, Gary Kasparov. In 2011, the public consciousness was captivated by a IBM computer competing against human opponents on the game show Jeopardy. Watson (named for the company first CEO) beat former winners Ken Jennings and Brett Rutter. What is astonishing about the computer's accomplishment is not just that it had the knowledge to answer the questions correctly, but it's ability to understand natural language, and in particular the type of "Trebek speak" that is characteristic of the show's host, Alex Trebek, which relies on double-entendre and word play. Watson is not making headlines with TV appearances, but it is not sitting idle. The program is learning more and it's expertise proving beneficial in a number of fields, notably medicine. It is now being used in health care to aid in diagnosis and recommends for treatment options. Watson's knowledge will soon be available as a companion toy for children. If you are tempted to find our futuristic ideas for language learning somewhat fantastical, consider what someone 25 years would have thought if you told them that soon there would a child's toy that you could have a real conversation with and would contain almost all knowledge known to mankind. Check out the video. Today's AI's is at work in Web search results, self-driving automobiles, speech recognition software, robot vacuum, recommendations for things you might like on Netflix, and facial and object detection in photographs. But the technology is advancing rapidly. Kevin Kelly from Wired.com identifies three recent breakthroughs that are propelling it forward:
|
|
References - hyperlinked to sources
Computer History Museum - Timeline of Computer History
A (Very) Brief History of Artificial Intelligence (Bruce G. Buchanan) - AI Magazine Magazine Volume Volume 26 26 Number Number 4 4 (2006)
Examples of Artificial Intelligence
The Three Breakthroughs That Have Finally Unleashed AI on the World - Wired.com
Computer History Museum - Timeline of Computer History
A (Very) Brief History of Artificial Intelligence (Bruce G. Buchanan) - AI Magazine Magazine Volume Volume 26 26 Number Number 4 4 (2006)
Examples of Artificial Intelligence
The Three Breakthroughs That Have Finally Unleashed AI on the World - Wired.com