So you've seen whats possible when you combine some circuit modifications and open social and web 2.0 applications. So what's next after that?

Well personally I would like to see Teddy get some mechanical modifications that give him the ability to walk around and maybe dance; or give teddy vision with some embedded cameras.

What I really would like to see, is Teddy become a learning err uh machine. Machine Learning is a growing sub-field of Artificial Intelligence, and focuses on extracting information from data automatically. Having ubiquitous access to data makes this a possibility, and leads to thoughts of how machines, such as Teddy, can access as much data as possible without having to store data in its own memory.

For example, if I wanted Teddy to "learn" as much about me and my friends as possible, I could program a database of all the information on Teddy's internal memory, which would require me to gather all the information about my friends, enter each entry into the database, and store it on Teddy. Now say one of my friends moves or changes some information about them, then I would have to in and manually change the information. This means I have to keep up with all my friends information and make the changes, which sort of defeats the purpose of Teddy keeping the information for me.

Well instead of embedding the information inside of Teddy, I link Teddy to each of my friends' data feed from facebook, myspace, or any other social media service that dynamically changes whenever my friends have new information. Now, Teddy does not have to store information about each friend, but only a link to each friend. As long as Teddy has access to the internet he can stay up-to-date with the latest information about my friends and keep me informed as well!

Now lets take that a step further and lets have Teddy perform a task for us based on our friends' information. For example, say we want Teddy to send a message to each of our friends when it is their birthday. We simply program Teddy to check for birthdays occurring on the current date, and have him automatically send a message to the person on our behalf. When Teddy sends the message, we will have him alert us of the persons birthday so we can follow-up and know who's birthday it was. We have thus transformed Teddy into a toy that not only helps us remember birthdays, but also performs a useful task for us.

As more web 2.0 applications become available and open, the possibilities continue to grow for machine learning and artificial intelligence. Even automobiles will benefit from advances in machine learning. Chrysler just announced that they plan to have in-car wifi in their 2009 lineup. Many find this to be a dangerous move by the Automotive giant, but I see this as a well overdue upgrade. The concerns are that drivers may get distracted by this feature (as if cell phones aren't enough). But from a technological perspective, this opens the door for automobiles to send and recieve real-time information about its performance, location, speed, route, etc. Automobiles could receive information about nearby cars, and make better routing decisions, or alert the driver of potential hazards ahead of time.

Going from a Teddy bear that speaks Twitter feeds, to one that provides a human-like emotional response or cars that let us know where the car in front of it is heading will not be an easy task. But as we continue to make our wireless and wire line networks more and more ubiquitous, only time will tell what great project is up next on the horizon.

Posted by Guru Lloyd | tagged: Teddy, Twitter, engineering