It's a question that's as old as Computer Science itself. Can we construct the perfect intelligent machine? While the perfect intelligent machine should more than just ace the Turing Test, this article at Technology Review on a Q/A session with Peter Norvig, Google's Director of Research, seems to suggest that our online search companion involves lot more Artificial Intelligence (AI) than meets the eye. Also, a pending patent suggests that Google may be targeting the TV audience very fiercely in the future.
The conclusion cannot be too far-fetched. Google has been investing resources heavily into Machine learning, natural language processing, and pattern recognition. Combine all this with the largest database of content and the result is monumental. In fact, Google's CEO, Eric Schmidt, has opined that the search gaint aims to target users so accurately that one day it may even suggest to you what job to take up (Financial Times).
Talking about targeting users, Ars Technica reports that a pending Google patent may take targeted advertising over TV to another dimension. Basically, the system will respond to users viewing pattern (by recording the sound emanating from the television) to pull up related Web pages on the fly. For example, you could be watching an episode of "Friends," and the system could push through pages referring to the outlets that sell the clothes worn by the cast or communities that discuss the characters.
More than personalization is the risk to user privacy. But provided that the increased personalized experience has its inherent insecurities, what is your take on this? Here's the poll: