The goal of voices is to write the framework software for creating chatbots whose intelligence evolves with the input from users who visit the chatroom in which they reside. Instead of blatantly attempting to imitate human intelligence, however, voices bots are programmed to constantly accept and intelligently organize incoming data. This gives them the capability to learn real-time from every day, average users — not just programmers.
Visit project website
voices's primary method for categorizing data is by tagging: the practice of using simple words or phrases – which is a method of describing that can easily be cross-referenced by computers – to describe other sorts of media that are not so easily understood by our digital desktop friends.
While it is not voices' goal to pass the Turing Test, the interaction between a voices chatbot and a user has been made to be as naturally human as possible. This has the strange side effect of appearing as though the bot is eavesdropping, as it does not necessarily care whether a user is addressing her when it decides to store information.