|
Biological Intelligence and its
Application to Large Companies |
|
|
|
|
Much
knowledge yet unused
A variety of scientific disciplines
are concerned with understanding and
modelling aspects of intelligent
behaviour:
- The neuro-sciences are dedicated
to the study of nerve cells,
their inter-connections and the
ways these connections can be
modified by learning processes.
- Cognition psychology looks at
algorithms transforming passive
data into case-sensitive
information and -at a higher
psychic level- the role symbols
play in the combination of such
information according to the
rules of logic.
- Psychology provides hints at how
the psyche of man adapts itself
to mastering a highly
unpredictable and complex
environment.
- The biology of evolution
contributes striking examples of
highly specialized and
well-adapted organisms emerging
without any intelligent
intervention necessary to the
feat.
- Genetics suggest a score of
mechanisms implemented by
biologically evolving organisms.
- The computer sciences are partly
successful in modelling some of
the results of the branches of
science mentioned above as to
produce quasi-intelligent results
(genetic algorithms, neural
networks, artificial
intelligence).
- The game theory
points at difficulties and
advisable strategies in dealing
with potentially hostile
competitors in an environment of
limited ressources.
- Ethology and the social sciences
observe and describe phenomena
displayed by small and large
groups of biological individuals
as they try and adapt to the
demands of their respective
environments.
- Mathematics and physics have
given birth to the the theory of
chaos and order, which covers the
border region between fickle
transcience and barren
inflexibility i. e. where
intelligent life forms are found.
- The economic sciences -often
without fully appreciating the
fact- deal with intrinsically
biological problems such as
companies competing for limited
ressources. Some of the
solutions, such as benchmarking,
conspicuously resemble biological
counterparts.
Over the past two decades, many
scientists have distilled from the
empirical and theoretical data amassed by
the classical sciences basic principles
of successful and intelligent behaviour.
Unifying concept such as complex adaptive
systems, "systems
sciences" and
"cybernetics" seem to be
particularly fruitful focii to this
process, rallying and merging
contributions from many different
disciplines.
The metaphor of a global organism has
gained new impetus with the continuous
emergence of ever more striking analogies
between the internet and nervous systems.
|
|
|
|
|
|
|
The idea
Large companies can be understood as
complex adaptive systems,
being capable of intelligently adapting
themselves to a constantly changing
environment. They are -at a certain level
of abstraction- comparable with
biological organisms. The question
arises, in how far one can not only
compare but actively transfer elements of
intelligence from biological organisms,
neural networks, genetic algorithms or
whole societies to large companies:
- Could a single human or a well
defined computer be treated as a
neuron, reducing many ingoing
signals to a few outgoing
signals?
- Could synaptic
plasticity be remodelled
within a future communication
software so as to alter the
probability with which
information from a certain source
can reach a certain recipient
under certain cirumstances?
- Could the process of
back-propagation, often used to
teach neural networks, be
imitated on a company scale by
allocating communication
resources according to the
participation of certain
communicators to succesful
decision and projects? Would this
give rise to a whole company
acting as one neural network?
- Could future browsers within an
intranet be equated to symbols
described by psychologists in the
way that they serve to collect
specific information, using
associations (links, key-words on
the internet)?
- In how far could the avowedly
important role emotions play in
influencing the data-processing
in our brains be copied by large
companies? What would a company
limbic-system look
like?
- Animal and human brains are
spatially highly differentiated,
thereby optimizing the use of the
limited lines of communication.
Should companies, too, try and
reassess their spatial structure
in order to minimize the demand
on data-transfer infrastructure?
- Can the various instruments
emerging under the name of
data-mining be understood in
analogy to the propensity of
human brains to try and find
causal relationships in whatever
data they can get hold of?
- What would be the effect of
successful companies not
expanding (thereby probably
discarding many of the qualities
by which they achieved success)
but duplicating themselves and
thereby their reasons for
success? Would this imitation of
biological organsims lead to a
more decidedly evolutionary
economic environment? Could large
investment-fonds and the
stock-exchange initiate or force
such a step on companies? Are
chain-companies pre-adapted to
this development? Could encoding
standards such as ISO 9000,
SAP-structures or clearly
written-down company profiles be
regarded as an incipient form of
DNA, serving as a blue-print when
a successful company is to be
duplicated?
|
|
|
|
|
|
|
The vision
Future companies make maximum use of
strategies of success provided by
biology. Their communication structure
will show distinct features of neural
intelligence, a single human or computer
being equatable to something on the level
of a neuron or even below.
A large population of similar companies
can be understood as a genetic
population. Success will be defined in
terms of reproduction (thereby sowing the
seeds of their success) rather than
growth. Genetic principles will be used
to optimize the adaption of the
population as a whole.
Any economic system redesigning its
companies along these principles will be
vastly superior to any other system.
|
|
|
|
|
|
|
A
first-step: a one-year assessment phase
The wide range of potential
contributors to these ideas and the
naturally limited understanding the
initiators of this scheme have of the
sciences adressed make it unadvisable to
look for partners at too early a stage.
Instead, the initiators suggest the
following step-by-step proceeding:
- IBH first of all creates a
vividly and distinctly modelled
phantasy-company, including many
features mentioned above. This
will basically serve to make the
vision communicable. Also,
current literature is gathered
and indexed in reference to the
various points of interests. All
results will be published on the
internet, together with the
procedure suggested below.
- Intellectual Property Rights and
other contractual issues will be
settled at this stage, using http://www.cordis.lu/ipr-helpdesk.
- Using the infrastructure of the
European community (http://www.cordis.lu/fp5/src/ncps.htm)
- The proposal will be announced
publicly, calling for potential
partners to come forward.
- IBH prepares a two-day workshop
at which the whole idea is to be
presented and discussed.
Participants should be recruited
from as many different
disciplines as possible,
preferably from computer,
biological and economic sciences.
Their number should be above 6
but not exceed 12. The aim of the
workshop is to formulate
fundamentally executable forms of
organization and the development
of programmable software,
transfering elements of
intelligent biological systems to
large companies.
- In retrospect, all participants
are asked to produce a written
assessment of the project as a
whole and a clear outline of
their specific contribution. IBH
will coordinate these
acitivities.
- All results will be documented
and published on the internet by
IBH. This will serve as a basis
for further evaluation as to the
continuation of the scheme as a
full project.
Money granted by the FET-option of the Fifth Framework
of the European Union should cover for
the period of the assessment phase:
- a one year salary of an academic
employee,
- the costs incurred by the
participants related to the
workshop (travelling,
accomodation),
- a pre-defined amount paid out to
participants willing to further
contribute to the scheme.
|
|
|
|
|
next page
|
|
|