I'm so Excited I could EXPLODE!

I'm posting this in the test forum so Bill can delete at his pleasure. ? 

The past few days for me have been absolutely incredible in terms of learning new information concerning my robot project.  I've learned so much that I need to really take a break and write some of this stuff up so I'll have it well-organized for future use.  I've made great progress in at least 3 areas of robotics.  At least as it has to do with what I am personally interested in achieving. 

The three main areas are,...

1. Learning about Speech Recognition Software in general.
2. Learning how to take FULL CONTROL of the Microsoft voices programmatically.
3. Gaining quite a bit of insight into my hopes and dreams of creating R.I.

Let's take these one at a time.

1. Learning about Speech Recognition Software in general.

After quite a bit of frustrated searching I finally discovered a fairly easy way to create a custom speech engine from scratch using Python.   I was concerned about this because I would eventually like for my speech engine to be totally independent of any copyright or propitiatory restrictions.   Just knowing that I will be able to program my very own speech engine from scratch using Python is extremely satisfying.  It won't be a program I could construct overnight.  But just knowing that it's possible and ultimately not even difficult to do has been an extremely pleasant thing to learn. ? 

So I want to make sure to document this well while it's fresh in my mind.  I might even make a video on it when I actually go to write the program.

2. Learning how to take FULL CONTROL of the Microsoft voices programmatically.

This is another thing that was quite difficult to learn via searches.  I really had to "pull Internet teeth" to discover the secrets.   But I finally found out how to do this.   I wanted my robot to sound like a young girl of say 4 or 6 years old.   I'm going to try my best to make my robot appear as a child.  A robot child to be sure.  She will have a wheeled base.  So there's no attempt to try to build a human child here. ? 

But, as crazy as this sounds I was concerned about finding a child-like voice for her.   Searching the Internet for child-like voices for speech synthesizers proved futile.  But then I learned how to modify and manipulate existing computer voices.   I was able to take a standard female voice synthesis and remodel it to sound like a little girl.  I learned how to change the pitch of the voice as well as the rate of speed she speaks.  And there are many other attributes I now have control over.   So I'm excited about this.

3. Gaining quite a bit of insight into my hopes and dreams of creating R.I.

Finally I decided to buckle down and bite-the-bullet as they say and start reading up on A.I.  Artificial Intelligence.  I found several website that offer up the "Best" books on A.I.  I even downloaded 6 of them, each close to 1000 pages long.  However, skimming through them I was gravely disappointed.  They are indeed all about A.I. (Artificial Intelligence).   None of them addressed the issues I was hoping to learn more about.   So I went back to re-read Marvin Minsky's book, "Society of Mind", and I am so glad I did.   Ideas just started pouring into my head as I re-read this book.   Keep in mind that I had read this book years ago, and studied it in some depth.  But re-reading it now has really opened up a flood-gate of new ideas.

Couple this with some psychology pages I've been reading about how human baby's thoughts and mental behaviors evolve over the first 4 to 8 years of life.  I know have some really good ideas for starting to try to create my own R.I.  Yep that's right. (Real Intelligence) ? 

No more of this artificial intelligence for me.   I'm going for the real thing this time around.  And now I have a really good way to measure my progress.  I'm going to measure it by how well my robot compares with a human baby at different ages.   It takes a human baby up to a full year just to be able to put together simple 3-word sentences.   Like "I fall down", or "I go school".  And supposedly these babies only have a vocabulary of 100 to 200 words at age 1.    I don't know how they can be sure how many words are floating around in a toddler's head, but 100 to 200 seems like an underestimate to me.  I don't  think they are giving the toddlers the credit they deserve.

Anyway if I can get a robot that can construct these kinds of meaningful sentences on its own in a meaningful context, then I can say that my robot is on the level of a 1-year-old baby.   So now I have a goal to shoot for, and if it takes a year to do it that will be just about right. ? 

So now I'm mentally exhausted from all this research.   I need to take a break, write it all up, and then start over again from a slightly higher level.   I have ideas of how to create this R.I. program and I would love to start programming.  But it's time to take a break and get back outside and work on tractors and logs some more.  Summer's going to be over and I won't have my winter prep done.  Then I'll be in trouble!

Caution: This post will self-destruct when Bill decides to delete the posts in the Test Forum.