After using Eliza and Alice, I can say that neither of them passed the Turing test. If the Turing test evaluates if an AI could pass as a human, then Eliza and Alice are very par from passing. They both had automated responses that didn't really pay attention to what the human said, but just copy and pasted words from human responses and repeated them back. The main problem with functionality is that they need better word detection software and more varied responses.
No comments:
Post a Comment