The idea that fully-fledged artificial intelligence (AI) — thinking machines that could mimic human intelligence precisely — would be realized by the year 2000 was driven by early breakthroughs in the field, such as the Logic Theorist program. The program successfully utilized a computer to solve logic problems through a virtual reasoning system that employed decision trees. However, since then the AI field has been one of deferred dreams, as many of the milestones predicted for the turn of the millennium have yet to come to pass.
"The reality of the engineering requirements [of AI] and what it really takes to make this work was much harder than anybody expected," says David Ferrucci, leader of the IBM Watson project team. He cites the development of chess-playing computers as an example of an advancement that created false hope about AI overall. The enormous difficulty computers have in emulating distinctly human abilities, such as communicating with people via natural language, demonstrates what a hard challenge AI creators face.
Among the latest AI achievements is Aberystwyth University's Adam, a robot that can make scientific discoveries using a method known as abduction. "Adam can . . . abduce hypotheses, and infer what would be efficient experiments to discriminate between different hypotheses, and whether there's evidence for them," says Aberystwyth's Ross King. "Then it can actually do the experiments using laboratory automation." However, King says the really complex problems involve humans interacting.
View Full Article
Abstracts Copyright © 2009 Information Inc., Bethesda, Maryland, USA
No entries found