"Firstly, Prolog is it was great at the deductive, expert system type of AI, but in 90's, a new generation of AI based on statistics (firstly, fuzzy logic, and then proper statistical reasoning, particularly Bayesian approaches) appeared and showed that truth and failure just don't cut it any more."

There are attempts to reconcile the two approaches, such as Markov Logic.

http://en.wikipedia.org/wiki/Markov_logic_network

I believe Statistical Relational Learning is the more general term for this idea.

http://en.wikipedia.org/wiki/Statistical_relational_learning