Why patomic theory
Instead of the popular view that a brain identifies objects by matching generalised
representations, it proposes brains match specific examples to identify the general.
Recognition, the key features of a brain, is very different to recall.
Representing real world objects and rapidly matching them is proving too complex.
Instead, storing multiple instances and linking them together in their relevant
location enables matching. For example, an image of a face stored 50 times is potentially
much simpler to match than a single instance.
Simply, the brain links together stored sensory patterns at their source and selects
stored motor patterns. It is not a general processing device, but a collection of
dedicated pattern-matching elements organised in a specific way.
Any brain theory needs to account for the observations made in the cognitive sciences.
For example, a human brain can recognise and react to complex situations in only
a quarter of a second. As the brain's materials operate in millisecond timeframes,
there are very few steps available for a brain to initiate coordinated muscle contractions.
Even the act of throwing a ball requires the brain to signal muscles at a rate too
fast for feedback. These observations show that a brain is sending a set of predefined
signals, rather than "somehow" processing motion equations.
Perhaps the most important aspect of a brain theory is its ability to incorporate
our evolutionary history. Stepping through the evolutionary tree, brains share many
common features which is why our ancestor's brains provide clues to ours.
Brain-damaged patients provide examples that our brain extends previous capabilities,
rather than replaces them. Human brains aren't uniquely different from other animal
brains suggesting many parallels can be drawn.
Although the conclusions forming patomic theory are counter-intuitive to most scientists,
it aligns with current observation about a brain capabilities. As with any new scientific
proposal, the revolutionary potential is yet to be fully tested.
Why look for a new model?
Despite more than fifty years of dedicated research into AI, there are no intelligent,
speaking robots on the planet. Something fundamental appears to be lacking in the
current approaches. What other explanation is there for our lack of progress? Human
beings always seem able to explain natural phenomena. We must be on the wrong track
because the question "How the brain works" remains unanswered.
Patomic theory explains the human capabilities and robotic requirements below:
A human brain
- can produce speech within 100 brain cell activations
- readily outperforms computers on real-world tasks, despite being
ten million times slower
- connects elements in a hierarchical, bidirectional manner
- controls walking on two legs with ease.
Today's robotic brain does not
- learn
- generalise from experience
- use and coordinate senses like vision, hearing, touch, taste, balance
or movement
- make decisions using emotions to determine the best or worse potential
outcome from actions
- make future plans
- recognise a human or other animal's face, in general
By understanding how to emulate these requirements, patomic theory provides a path
for researchers and cognitive science professionals.
More information
Through the exploration of pattern-matching technology, Patomic theory promises
to contribute to the quality of life for patients of neuroscience, linguistics and
psychology. It also enables new approaches to philosophy and artificial intelligence
which will also lead us to further insights about ourselves.