Google is also close to natural language understanding; it can answer simple questions up to a few words.
Not really. Given a simple short question, it can present a human with text that a human can recognize as containing the answer (among other things). That is pretty different from answering the question. PowerSet, or some other scalable NL search method based on a fusion of rule-based and statistical parsing, may end up being able to answer simple short questions based on the massive knowledge on the Web though. I do think this is do-able with pre-AGI technology, but Google isn't gonna get there with its current, non-NL-parsing-based methodology. (However, they have hired plenty of statistical NL parsing experts, so they may well beat PowerSet there ... or acquire PowerSet, etc.) Ben G ----- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415&user_secret=fabd7936
