The evening sessions tonight started off with some interesting “open
mic” Q&A … some of the highschool students from the audience
spoke about their perspectives … it was interesting to hear.

Jerry got up to start talking about Brave New Virtual Worlds … and
jumped right into some of the interesting things going on in Second Life. 
He talked about how there are now all sorts of real-world to
virtual-world interactions.  One example was a Cancer Society fund
raiser where Second Life characters walked in Second Life to raise
money in the real world.  In addition, Jerry even holds regular
Future Salon meetings within Second Life where other Second Life
avatars come to participate and listen to virtual presenters.

He briefly talked about the Croquet Project … something that will be
in a talk tomorrow.  Another virtual world project.  I
downloaded the Second Life client software and will sign up for a free
account … I want to see what this is about.  Hmmmm … my first
virtual identity in a 3D virtual world.

The next breakout was supposed to be a video conference with Robert
Hecht-Nielsen, but something didn’t work out … so they chose to play
a DVD of his presentation. Still interesting. The best part
was the term “Confabulation Theory”! Yeah … confabulation is a word that is in the dictionary.

His presentation explained how four key concepts form the foundation of cognition.

He started by proposing that cortical modules exist in the brain, and each of these is responsible for describing one attribute of Objects in the Mental Universe. These cortical modules are made up of groupings of neurons in the form of symbols that we learn at an early age. Each symbol is made up of ~60 neurons. Symbols could be names, colors,
or any other attribute of an object. As numerous cortical modules
are receiving sensory input, confabulation occurs … triggering
behavior.

Again, he emphasized that it is the quantity of knowledge that allows
for confabulation to occur. Billions of knowledge links.
One third of our life or more is spent learning the foundations.
Language is the essential core
of cognition.  As we grow up we develop a rich set of symbols, and
interconnections between symbols – between cortical modules – by
hearing words combined with other sensory experiences.

There is more information about his theories in this article about his new theory of cognition.

After lunch at Accelerating Change 2005
we progressed into a series of breakout
sessions. The first of of my choices was tough, but I went to a
session titled Shrinking the Planet given by:

Peter Barrett, Microsoft IPTV
Scott Rafer,
Feedster

Peter started off commenting on the growth of bandwidth to the home,
and also the improvements in codec technologies. Both of these
are combining to create an environment for on-demand IP delivered video
content. This environment is quickly becoming the foundation for
supporting the long tail
concepts for video. He showed some demos of IPTV, and talked
about the metadata that will be exposed in the TV guide – like how many
people are watching a particular show, or even specific people you know
like your family and friends. He said that the project is based
on the DVR concepts, and that one of the initial partners is SBC with
their Project Lightspeed.

Scott began his talk with some thoughts that are similar to mine …
that AI might just come from millions of interconnected humans.
He referenced Dodgeball, and expanded on this theme with various potential mash-ups of various datasources. His slide titled Humans Will Be My AI
gave a number of examples where large amounts of data – aggregated by
humans – become the valuable source of metaperspectives. It was
almost funny how many people were uncomfortable during the Q&A with
his perspective of AI not necessarily being something that is
created on a computer from some lab. Maybe one day … but sooner
then that I believe that humans are quickly becoming the “sensors and
actuators” at the edges of a growing AI …