Artificial intelligence (AI) has become a hot topic again thanks to Watson from IBM. Automatic translation has often been considered the Holy Grail of AI. Countless knights from the world of AI have been searching for it. But literary text translation and poetry in particular have so far only made computers look foolish. The Grail of AI has remained elusive as it can only be found by a “hero who must prove himself worthy to be in its presence. He must grow spiritually and mentally before he can locate it”.
But things are changing. IBM’s Watson supercomputer has won a difficult quiz show. It is not the first computer from IBM to achieve a spectacular victory over man; 14 years ago Deep Blue defeated former world champion Gary Kasparov at chess. For those familiar with Jeopardy, it is clear that this is an outstanding feat. Most artificial intelligence systems experience difficulties with their “knowledge of the world”, proving that they’re unworldly. Another major problem is the correct interpretation of ambiguous questions. Both problems have now been tackled successfully by Watson.
So, has Watson found the Grail? Maybe. Even though it’s only a quiz show, it’s one that requires a lot of interpretation, perhaps comprehension, and “knowledge of the world”. All of this is essential for good translations. So it seems that the AI Grail is now within reach. Watson has shown himself worthy to be in its presence. And now for the final stage: can it grow spiritually and mentally?