CABot5
- The current grant runs out at the end of October.
- However, we've already submitted a follow on grant for
CABot5.
- It's a two prototype system with CABot4 and 5.
- The main issue is to learn better CAs (CA3).
- We also propose procedural learning including rules with variables.
- It will be a conversational agent so the system will also
produce text.
- Vision will include foveation and parallax.
- If we don't get the grant, I'll still continue on with the CA3
stuff, but we could also go different ways with this.
- Beyond, or in addition to CABot5, we can look into a range
of other things including attention, speech recognition,
and improved environmental understaing.
- I don't expect this research to end in my life time, but if
we do it all, we will have an AI.
