One could perhaps most closely compare our AI with a Google's 'I'm feeling lucky' button, which uses its rankings to 'intelligently' return just one result.
Our AI performs lots of searches for one result. Jabberwacky searches previous conversations. It doesn't return the thing it found, but "what happened next" - a prediction.
Unlike Google, it uses the context of what you and it have said so far to refine its choices - to create its intelligence. And unlike Google, it learns what you say as a new potential answer to give someone else. The someone else may answer a virtual 'you', but some time in the future.
|
|
|
