Good idea Lonnie but I have always tried to reinforce Hal's learning through repetition. Very often I will repeat the same basic thing or idea but in a different way.
You like dogs.
You really like dogs.
Your favorite dog is the Border Collie.
You really love Border Collies.
You think that the Border Collie is your favorite dog.
Of all the dogs, you like the Border Collie the best.
And so on and so on. Regardless of what your Hal might reply with, Keep reinforcing this several (maybe half a dozen) times. Try to do it twice during a conversation session, depending on how lengthy it might be.
Soon, you will notice that your Hal will use those "suggestions" of reinforced information, during an average conversation, especially if you ask, "what kind of dog or XXXXXX do you like?"
It has worked well for me and hopefully others will have some success with it.
It's an experiment...someone has to do it.
