|
For scientists, journalists and politicians and anybody interested
Next to thinking, “information
management” is an extremely important function of the brain.
However, publications about the brain seem to lack fundamental
information management/architecture aspects as seen in
computers. Here are
some of those fundamental aspects:
-
Education, travel, TV and the
Internet create the need for the human brain to process and
store dramatically more information than ever before. On the
other hand, there must be “pathways” within the brain to
connect the different parts and data/patterns. As with
computers, this alone sets the stage for capacity
bottlenecks to develop.
-
Capacity bottlenecks can be
extremely difficult to find in computers and can express
themselves in a variety of symptoms.
-
When it
comes to capacity bottlenecks, a small difference can have a
huge impact. This may be why the results of previous
scientific studies vary and conflict with each other.
-
The speed at which neurons
fire is slow. Thought-to-muscle activation needs to happen
within only 100 steps. (See Hawkins’ 100-step rule [1].)
-
Parallel processing appears to
be seen as the way to overcome the slow speed; however,
parallel processing only works well to the point at which
data must be shared. From this point on, locking mechanisms
must be used so one process cannot overwrite the
data/patterns of another process. With more processes
working in parallel, this gets extremely complex and
conflicts with the speed at which neurons can fire.
-
The technique of “addressing”
is needed for processes to locate data/patterns in a
fraction of a second. This is a complex technique. How could
nature have developed addressing?
-
Moving data/patterns around
adds another level of complexity.
-
An increasing number of
experts (for example, Jeff Hawkins [1]) provide
reasoning as to why complex matters such as making robots
catch a ball and performing advanced language translations
cannot be done through traditional programming techniques.
They are simply too complex.
This leads to the conclusion
that information management within the brain may be the opposite of complex, instead using a rather simple
architecture through which these complications can be
avoided.
Computer-based neural networks
are simple and loosely based on real nervous systems. However, they didn't bring
the breakthroughs hoped for as they missed structures such
as hierarchy [1].
The
information management model outlined in Surprise Treatment
for ADHD, Dyslexia, Headaches and other Conditions - It's All
About Information Management is rather simple. The
architectural issues and conflicts listed above are non-issues
in this model.
The model is a neural network
model. It is self-organizing and dynamic (dynamic: neural
structures are re-used as their content is 'forgotten'; re-use
can be for different things). It has functionality that should not exist with
traditional neural networks. It's this additional functionality,
combined with connectivity challenges in real brains,
through which many things become explainable.
[1] On Intelligence. Jeff Hawkins, Sandra Blakeslee; 2004; Times
Books, Henry Holt and Company, ISBN 0-8050-7456-2 |
