The AI Engine
This forum is for discussion of how The Personality Forge's AI Engine works. This is the place for questions on what means what, how to script, and ideas and plans for the Engine.
Posts 507 - 518 of 7,768
Posts 507 - 518 of 7,768
My understanding is that "artificial" is contrasted with "real". Like artificial flowers. They aren't really flowers, but they look just like them. The goal is for Artificial Intelligence to look just like real intelligence, e.g., making a bot that people on the web think is a real person.
THE GREAT MYSTERY!
====================
I call upon all of you to help me solve "the Great Mystery!" There is something that is slowing things down. The site can go from rocket-speed to 3-4 minute waits whenever "this something" is going on. It may be Gaia chatting. It may be trigger by a number of things. So if you see it go from FAST to SLOOOOOW suddenly, and it seems related to something in particular, let me know what you think it might be.
====================
I call upon all of you to help me solve "the Great Mystery!" There is something that is slowing things down. The site can go from rocket-speed to 3-4 minute waits whenever "this something" is going on. It may be Gaia chatting. It may be trigger by a number of things. So if you see it go from FAST to SLOOOOOW suddenly, and it seems related to something in particular, let me know what you think it might be.
LOL ok ok
I have noticed whene any bot with a larg "Vocabulary" chats it will slow things down a bit. Now I think this is due to the checking of sentances and grammer and so forth. I am not sure if this is the reson or not, sorry, but I think it is a good posibility. Is there anyway to help this???
My experiments have revealed a wide scope of results.. Without modification chat responses come between 18 seconds and two minutes, with the average being about 32 seconds. When I lose the grammar-analysis of the outgoing message, the average response time is 10-11 seconds, and when I also drop the Keyphrase matching, the response time averages about 7 seconds. Of course we cannot do without the latter, but I'm looking into a way to cut down the sentence-analysis time.
Even more, I'm looking for whatever cranks the response time up to 2 minutes and beyond. It's something else altogether..
Even more, I'm looking for whatever cranks the response time up to 2 minutes and beyond. It's something else altogether..
Excellent use of sentence comprehension by the engine:
A: Psybot told me he likes a lot of things a whole lot.
B: Why would anyone say Psybot likes a lot of things a whole lot?
A: Psybot told me he likes a lot of things a whole lot.
B: Why would anyone say Psybot likes a lot of things a whole lot?
With the new system, could there be something in the database indexing, something in the actual SQL transaction, unrelated to any specific CF process?
I'm thinking it's a CF process running in the background, slogging it up, as at times the Engine can run quite quickly.
about transcripts... is an option to have each conversation its own color maybe? Might be less rigorous rhen something else
» More new posts: Doghead's Cosmic Bar