Imagine a baseball game. A pitcher throws the ball. An event in time. What happened? Did the batter swing? Did he connect with the ball? Where did the ball go? Did it leave the infield? Was it caught? Was it a home run? Did the runner get on base? Was it an out? Was it the third out?
So many events? Which raise questions. And the answer raises new questions. Which spawns more questions.
All these events can be captured. And many of them are. To create stats. And there are paid professionals who mine this data for a living. And their demand is increasing.
Each event must be captured and stored in an application connected to a database. And the process for doing so is most likely a person. And he or she views the event, records it, for later consumption. Once the event is over, it's gone. So when the deep mining algorithms used to process the data has a one directional inflow of information, a series of events in chronological order. The process is not two way, in that as the system is fed the information, it has the ability to assess, process and ask questions in real time.
It's like explaining a song to a deaf person. Or visualizing a sunset to a blind person. Sure they understand what's happening, but they can't feel the emotions of a Mozart Symphony or a brilliant orange sun drifting beneath the horizon. There is no connection of live feedback, it's difficult to feel the emotions, which bring up memories, which spawn new emotions in an endless cycle.
It's simply a one directional information feed processed in real time or at a later date. And because of this, the picture is limited, it's not holistic or interactive, like the hard of hearing person not fully hearing the music, they only get a subset of information which is limited, which doesn't have the two way connection.
I think that's the next phase for true Artificial Intelligence. Open the awareness of information in real time to create a link so a two way communication can be established to allow a more dynamic interpretation of peripheral input for greater interpretation. Like viewing the baseball game, a two way feedback system could allow the system to kick off multiple queries based on current events, and depending on the next result, it spawn additional queries, some looking for past events to predict potential events, statistically weighted and then verified for accuracy in real time, which gets stored in the brain of the machine for later retrieval and analysis.
Computers lack an awareness, because of lack of interaction with their environment, in real time, because it's a one way communication. By allowing a two way communication to occur, the AI system could run more jobs in parallel to predict what's happening, and adjust it's behavior accordingly. What the computer really needs is to take some LSD to expand it's awareness, at this point, it's just to "square".
I signed up for the Hortonworks Certified Associate exam last Thursday. Figured if I sign up, I'd have to take the test. And if I tak...
This blog post is in no way an attempt to steal other people's work. It's basically an conglomeration of notes from research I did...
Saw a post today on Twitter, " Microsoft releases CNTK, its open source deep learning toolkit, on GitHub " This is big news. Be...