Connectionists: Brain Principles Manifesto (Version 2.0)
Juyang Weng
weng at cse.msu.edu
Tue Feb 24 12:27:03 EST 2015
Dear Colleagues:
I would like to thank all those who gave comments, some giving multiple
times. This is a version 2.0 as it explicitly includes some brain
mechanisms although it should not include those mechanisms that laymen
do not understand from just reading this document.
Some people said that he needs to read multiple times. Each time gains
more understanding. The information is very dense in this short document.
Those who are not on the BMI mailing list please subscribe to the list
to get further information. Simply do an internet search to get the
list site.
----
The Brain Principles Manifesto
(Draft Version 2.0)
Feb. 24, 2015
Historically, public acceptance of science was slow.For example, Charles
Darwin waited about 20 years (from the 1830s to 1858) to publish his
theory of evolution for fear of public reaction.It took about 20 years
(by the 1870s) the scientific community and much of the general public
had accepted evolution as a fact. Of course, the debate on evolution
still goes on today.
Is the public acceptance of sciencefaster in modern days?Not necessarily
so, even though we have now better and faster means to communicate.The
primary reason is still the same but much more severe --- the remaining
open scientific problems are more complex and the required knowledge to
convincingly understand goes beyond any single person.
For instance, network-like brain computation --- connectionist
computation --- has been long doubted and ignored by industry.Kunihiko
Fukushima introduced /Convolutional deep/ networks by at least
1980.Weng, Ahuja and Huang published /Max-pooling/ in deep fully
automatic learning networks by 1992.However, Apple, Baidu, Google,
Microsoft, Samsung, and other major related companies did not show
considerable interest till after 2012.That is a delay of about 20
years.The two techniques above are not very difficult to understand.
However, these two suddenly hot techniques have already been proved
obsolete by the discoveries of more fundamental working principles of
the brain.
Industrial and academic interests have been keen on a combination of two
things --- easily understandable but superficial tests and which
companies are involved.However, the newly known brain principles have
told us that the ways to conduct such tests will give only vanishing
gains that do not lead to a realistic zero error rate, regardless how
many more images can be added to the training sets and how long the
Moore’s Law can continue. Why?This is because all these static training
sets misled the participants by making it impractical to conduct serious
autonomous object segmentation that our human babies learn to perform
through interaction with the real physical world.Do our industry and
public need another 20 years?Or more?
Oct. 2011a highly respected multi-disciplinary professor kindly wrote:
“I tell these students that they can work on brains and do good science,
or work on robots and do good engineering.But if they try to do both at
once, the result will be neither good science nor good engineering.”How
long does it take for the industry and public to accept that that
pessimisticview of the brain was no longer true even then?
The brain principles that have already been discovered would bring
fundamental changes in the way humans live, human countries and
societies are organized, and the way humans treat one another.The
following questions point to some concrete fundamental changes that
benefit all humans. However, conventionally, scientists in natural
sciences do not address politics.Albert Einstein and Norm Chomsky are
among exceptions.
The brain of anybody, regardless of his education and experience, is
fundamentally short sighted, in both space and time, determined by the
known brain principles.Prof. Jonathan Haidt documented well such
shortsightedness in his book “/The Righteous Mind: Why Good People Are
Divided by Politics and Religion/
<http://www.amazon.com/The-Righteous-Mind-Politics-Religion/dp/0307455777>”,
although not in terms of brain computation.
In terms of brain computation, the precise circuits in your brain
self-wire beautifullyaccording to your real-time experience (the genome
only regulates) and their various invariance properties for abstraction
also largely depend on experience.Serotonin (e.g., caused by threats),
dopamine (e.g., caused praises) and other neural transmitters quickly
change the way these delicate circuits work but you feel everything
inside the brain is normal. Therefore, you make mistakes but you still
feel normal in the brain. Everybody is like that, including the
politicians in the questions below.
Surprisingly, to understand how the brain works requires a sophisticated
automata theory in computer science. See J. Weng, /Brain //as//an
Emergent Finite Automaton: A Theory and Three Theorems/
<http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CCAQFjAA&url=http%3A%2F%2Fwww.scirp.org%2Fjournal%2FPaperDownload.aspx%3FpaperID%3D53728&ei=21rnVP7fB4KZyAT-noH4Cg&usg=AFQjCNEzHz-YDhivI1esDtgKg84SEx4RuQ&bvm=bv.86475890,d.aWw&cad>,
/IJIS/, 2015, which proposed the following brain principles:
1.The developmental (genome-like) program uses fully emergent,
task-nonspecific internal representations of the brain-like Network (DN)
and therefore, the DN is of general-purpose --- can learn any
body-capable tasks, at least in principle.
2.The sensory area allows naturally sensed images of cluttered scenes in
which many objects mix, instead of an encoding that fits only what a
Turing Machine can work with. A teacher does not segment objects as how
children learn.
3.The motor area allows subareas where each subarea represents either
declarative knowledge (e.g., abstract concepts such as location, type,
scale, etc.) or non-declarative knowledge (e.g., driving a car or riding
a bicycle).
4.Every area in the “brain” DN emerges (not statically exist) uses a
unified area function that does not have local minima in its high
dimensional, nonlinear, and non-iterative approximation in its feature
development and adaptation.
5.The “brain” DN learns incrementally—taking one-pair of sensory pattern
and motor pattern at a time to update the network and discarding the
pair immediately after. E.g., the real brain has only one pair of retina.
6.The “brain” DN is always optimal --- every update of DN realizes the
maximum likelihood estimate of the “brain”, conditioned on the limited
computational resources in the “brain” and the limited learning
experience in its “life time”.
The logic completeness of the brain DN is (partially, not all)
understood by a universal Turing Machine which is like our modern-day
computer, in principle.This automaton brain model proposes that each
brain is an automaton, but also very different from all traditional
symbolic automata because it programs itself --- emergent.No traditional
automata can program themselves in the sense of Turing Machine but a
brain automaton does.
The automaton brain model predicted that neural circuits precisely
record the statistics of experience, roughly consistent with neural
anatomy (e.g., Felleman & Van Essen, /Cerebral Cortex/, 1991). In
particular, the model predicted that “shifting attention between
`humans’ and `vehicles’ dramatically changes brain representation of all
categories” (J. Gallant et al. /Nature Neuroscience/, 2013) and that
human attention “can regulate the activity of their neurons in the
medial temporal lobe” (C. Koch et al. /Nature/, 2010).The model raised
questions to claims that neurons encode exclusively sensory information
like the “place” cells in the work of 2014 Nobel Prize in Physiology or
Medicine instead of a combination of both bottom-up (e.g., place) and
top-down context (e.g., goal) as reported by Koch et al. and Gallant et
al. and theoretically predicted by the automaton brain model.
Unfortunately, the automaton brain model implies that all
neuroscientists and neural network researchers are unable to understand
the brain of their studies without a rigorous training in automata
theory. For example, traditional models for nervous systems and neural
networks focus on pattern recognition and do not have the capabilities
of a grounded symbol system (e.g., “rulefully combining and
recombining,” Stevan Harnad, /Physica D,/ 1990).Automata theory deals
with such capabilities. Does this new knowledge stun our students and
researchers or guide them so their time is better spent?
Understanding brain automata would enable us to see answers to a wide
variety of important questions, some of which are raised below. We do
not provide yes/no answers here, only raise questions.The automaton
brain model predicts that there is no absolute right or wrong in any
brain but its environmental experiences wire and rewire the brain.
In each of the following social science questions, we have two sides,
Side A with more knowledge due to the more open political and
ideological environment and Side B with less knowledge due to less open
political and ideological environment, but all have normal human brains
that should be respected.Is it more scientifically productive for Side A
to approach Side B friendly with caring heart so that Side B is not
threatened and not listening?We want them to rewire for good instead of
for bad, do we?
How can our industry and pubic understand that the door for a great
opportunity that has opened up for them?How can they see the economical
outlooks that this opportunity brings with it?
How should our educational system change to prepare our many bright
minds for the new brain age?Has our government been prompt to properly
respond to this modern call from the nature?
How should our young generationact for to this new opportunity that is
unfolding before their eyes?Is a currently narrowly defined academic
degree sufficient for their career?
Is it consistent with the U.S. people’s interest for the respected Mr.
Barack Obama to have authorized the bombing of ISIS, sanctioned Russia
because of what happened in Ukraine, rejected conversations with North
Korea for what the respected Mr. Kim Jong-un did, increased extra tax on
Americans who create many jobs, and planed to tax Americans’ overseas
ventures which encourages them to drop U.S. Citizenship? Shortsighted?
The same ISIS bombing question goes to the respected Mr. François
Hollande.What is the relationship between the armed attacks on the
weekly Charllie Hebdo and the French ISIS bombing that killed many more
innocent civilians as well as racial discrimination existing in France?
Is it consistent with the Chinese people’s interest for the respected
Mr. Jinping Xi to conduct anti-graft struggle using the Communist Party
rules without the due process of the Chinese legal system and to bicker
about islands with China’s neighbors like Japan, Vietnam, and
Philippines that negatively affected economy and Chinese visitors’
overseas safety?
Is it consistent with the Israelis people’s interest for the respected
Mr. Benjamin Netanyahu to take his current approach to Israel’s Arab
neighbors?
How should all government officials take advantage of the new knowledge
about their own brains?Should people in every country require them to
learn brain theory and correct their feel-normal mistakes?
We are from all walks of life and from all regions of the world.At
present, we do not understand the scientific underpinnings of the
material in this Manifesto, just like the public of Darwin’s
time.However, these issues are relevant to the future of our nations and
our lives.We declare to form the Brain Principles Society, in order to
promote human communication and understanding of brain principles and
their implications to human societies so as to improve the quality of
life for all human beings on this planet.There is a lack of society that
regards social sciences as part of brain science and considers automata
theory to be relevant to brain science and social sciences. However, we
are all governed by the same set of brain principles.
----
Please give your future comments, to the entire list or to me
individually. All past comments have helped shaping this document.
Best regards,
-John
--
--
Juyang (John) Weng, Professor
Department of Computer Science and Engineering
MSU Cognitive Science Program and MSU Neuroscience Program
428 S Shaw Ln Rm 3115
Michigan State University
East Lansing, MI 48824 USA
Tel: 517-353-4388
Fax: 517-432-1061
Email: weng at cse.msu.edu
URL: http://www.cse.msu.edu/~weng/
----------------------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20150224/c7b9e48d/attachment.html>
More information about the Connectionists
mailing list