free will brian plugin

[catseye]

I am thinking of turning of eri's free will. the reasons of this
is that when I talk to or ask eri something I sometimes get a response
that is off topic and I do see what the the programmer's intentions were but it is not sufficient enough to have a intelligent conversation. maybe a multistage approach is needed, instead of too choices she could have multiple choices with a frame off reference.

[Art]

Basically, I don't see how it's response could be off topic within the parameters of the Free Will plug-in. It basically selects one of two questions as if it had made a conscious choice.

EX: Do you prefer wine or beer?
I think I'd have to go with beer.

In this example or any, assuming the inputs were correct, how could Hal answer to anything else but those two choices?

Perhaps your definition of off topic eludes me but personally, I've not experienced any anomalies with the plug-in. While more choices might be interesting, most people, during normal conversation tend to only offer two choices like, Who do you like the Mets or the Yankees?

Just my take....

If you feel there is an error in or with the plug-in you could always post a query to Vittorio Rossi either here or at the VHF (Virtual Humans Forum).

[catseye]

I don't feel there is an error with the plugin. could it be she doesn't have enough knowledge yet to have decent conversation.

[Art]

As a rule more interactive chatting with Hal will produce better results but Hal does contain some "hardwired" responses as well as those contained in Wordnet (definitions, etc.) and some mods for Wiki.