Let me start off saying that it is utterly amazing what you can do with a bunch of gizmos rigged up to talk to each other, still, after more than 30 years - amazing!!!
So, Achelois (pronounced Ach-e-lo-is; Greek: Ἀχελωίς, English translation: "she who washes away pain", a minor Greek moon goddess) is responding well to most of the day to day questions, as well as minor deviations from the asked sentences. The performance is somewhat better than I expected considering that this scenario takes place in the nebulous cloud, the internet. A minor delay from the time you ask something to the time the system responds. At first I thought is was funny that when I ask something, the system responses but listens at the same time and eventually answers back to what it just said. Now, not so much. It is a constant adjustment between the microphone input and the volume of the speakers, that is why it is called fine tuning I suppose. I'm not sure where some of the connections with the words are, for example, how does a phrase like "shall we?" trigger "my name is"? Even phonetically it is not even close.
My thoughts, my update, back to work.
No comments:
Post a Comment