Transcript Groups - Department of Computer Science

Introduction to Complex Systems:
How to think like nature
Organizations: emergence and innovation
Russ Abbott
Sr. Engr. Spec.
Rotn to CCAE
310-336-1398
[email protected]
 1998-2007. The Aerospace Corporation. All Rights Reserved.
1
Flocking
• Craig Reynolds wrote the first flocking program
two decades ago: http://www.red3d.com/cwr/boids.
• Here’s a good current interactive version:
http://www.lalena.com/AI/Flock/
• A soccer game based on “forces.”
– Download, execute.
– After it starts, click Console tab
and reduce speed to 0.025.
2
Group/system-level emergence
• Both the termite and ant models illustrate emergence (and multiscalarity).
• In both cases, individual, local, low-level rules and interactions
produce “emergent” higher level results.
– The wood chips were gathered into a single pile.
– The food was brought to the nest.
• Emergence in ant and termite colonies may seem different from
emergence in E. coli following a nutrient gradient because we see ant
and termite colonies as groups of agents and E. coli as a single entity.
• But emergence as a phenomenon is the same. In both cases we can
explain the design of the system, i.e., how the system works. In the
ant/termite examples, the colony is the system. In the case of E. coli,
the organism is the system.
In Evolution for Everyone, David Sloan Wilson
argues that all biological and social elements are
best understood as both groups and entities.
You and I are each (a) entities and (b) cell colonies.
http://evolution.binghamton.edu/dswilson/
3
Breeding groups/teams/systems
Evolutionary processes are
fundamental to complex systems
Traditional evolutionary theory says there is no such
thing as group selection, only individual selection.
Bill Muir (Purdue) demonstrated that was wrong.
• Chickens are fiercely competitive for food and water.
http://www.ansc.purdue.edu/faculty/muir_r.htm
• Commercial birds are beak-trimmed to reduce
cannibalization.
• Breeding individual chickens to yield more eggs
compounds the problem. Chickens that produce
more eggs are more competitive.
• Instead Muir bred chickens by groups.
• At the end of the experiment Muir's birds' mortality rate was 1/20
that of the control group. His chickens produced three percent
more eggs per chicken and (because of the reduced mortality)
45% more eggs per group.
Wikipedia commons
4
Wilson on groups
Moral systems are interlocking sets of values, practices,
institutions, and evolved psychological mechanisms that
work together to suppress or regulate selfishness and
make social life possible. —Jonathan Haidt
• What holds for chickens holds for other groups as well: teams, military units,
corporations, religious communities, cultures, tribes, countries.
• Successful groups are those that minimize within-group conflict and organize
to succeed at between-group conflict.
• Groups with mechanisms for working together can often accomplish far more
(emergence) than the sum of the individuals working separately.
– E.g., most corporations, military organizations, etc.
• But if a group good is also an individual good (e.g., money, security), the
group must have mechanisms to limit cheating (free-ridership).
• Group traits (although they are carried as rules by individuals) evolve
because they benefit the group. (E.g., insect behavior.)
• Group selection (not just individual selection) now accepted as valid.
• These traits may be transmitted genetically (by DNA). They may also be
transmitted culturally (by training/parenting/indoctrination/mentoring/…).
– Human groups are much more complex because it’s not all built-in.
5
When cells reach the point where they divide
constantly, they are cancer cells. Instead multicellular organisms use a seemingly inefficient
process to replace lost cells.
An organ such as the skin calls upon skin-specific
stem cells to produce intermediate cells that in
turn produce skin cells. Although great at their
job, the new skin cells are evolutionary dead ends.
They cannot reproduce.
Losing the ability to reproduce was part of the
evolutionary path single-celled organisms had to
take to become multi-cellular.
What was in it for the single cells? They got to be
part of something more powerful. Something that
was hard to eat and good at eating other things.
Stem cells
instead of cancer
John W. Pepper, University of Arizona
Organisms are just a bunch of cells. If you
understand the conditions under which they
cooperate, you can understand the conditions
under which cooperation breaks down. Cancer is a
breakdown of cooperation.
If cells reproduce by simply
making carbon-copies of
themselves, their descendants
are more likely to accumulate
mutations. Suppressing
mutations that might fuel
uncontrolled growth of cells
would be particularly
important for larger organisms
that had long lives
Animal Cell Differentiation Patterns Suppress Somatic Evolution ,PLoS Computational Biology Vol. 3, No. 12, (12/2007)
6
We’re smart because we are “programmable,”
i.e., able to learn—both information and norms
As humans we’re successful because we’re smart.
We’re smart because we operate in complex groups.
We can operate in complex groups because we
have strong reciprocity.
Next slide
We both share and are willing to punish non-sharers.
Take bees.
You always think of the hive as the big social
collective. Not true. Workers often try to lay eggs,
even though only the queen is supposed to lay eggs.
If workers lay eggs, other workers run around, eat the
eggs, and then punish the workers that laid the eggs.
Wherever you find cooperation, you’ll also find
punishment. Think of your own body.
Each cell has its own self-interest to multiply. Why
don’t they go berserk (cancer)? How do you get cells
to cooperate? You punish cells that don’t cooperate.
Herbert Gintis
Clearly fundamental. How
are we autonomous?
Socialization: norm internalization.
There's no such thing in biology,
economics, political science, or
anthropology.
Humans can want things even
when they are costly to ourselves
because we were socialized to
want them
to be fair, to share, to help your
group, to be patriotic, to be honest,
to be trustworthy, to be cheerful.
What does it mean to say that we
can learn?
The word may sound cold and
robotic, but it means that we are
“programmable,” i.e., capable of
internalizing new skills and ideas.
Socialization is a form of learning.
7
Homo economicus vs. strong reciprocity
Homo economicus: individual selection
• Agents care only about the outcome of an
economic interaction and not about the process
through which this outcome is attained (e.g.,
bargaining, coercion, chance, voluntary
transfer).
• Agents care only about what they personally
gain and lose through an interaction and not
what other agents gain or lose (or the nature of
these other agents’ intentions).
• Except for sacrifice on behalf of kin, what
appears to be altruism (personal sacrifice on
behalf of others) is really just long-run material
self-interest.
• Ethics, morality, human conduct, and the
human psyche are to be understood only if
societies are seen as collections of individuals
seeking their own self-interest.
Strong reciprocity: group selection
• A predisposition to cooperate with others,
and to punish (at personal cost, if
necessary) those who violate the norms
of cooperation
– even when it is implausible to
expect that these costs will be
recovered at a later date.
• Strong reciprocators are conditional
cooperators
– They behave altruistically as long as
others are doing so as well.
• and altruistic punishers
– They apply sanctions to those who
behave unfairly according to the
prevalent norms of cooperation.
Moral Sentiments and Material Interests: The Foundations of Cooperation in Economic Life
Herbert Gintis, Samuel Bowles, Robert T. Boyd and Ernst Fehr (eds), MIT Press, 2005
8
Experimental “games”
C
D
• Prisoner’s Dilemma.
C
3/3
0/5
– One shot. Defect is the only rational strategy.
D
5/0
1/1
– Iterated.
• Tit-for-tat: Cooperate initially and then copy the other guy.
• Pavlov: repeat on success; change on failure. (More robust.)
A far from equilibrium system. New energy is supplied “for free.”
• Ultimatum Game. Proposer must offer to divide $100—e.g., from TAI.
Responder either accepts the proposed division or rejects it—in which
case neither gets anything.
– Only rational strategy: proposer offers as little as possible; responder
always accepts.
– Real experiments (world-wide). Responder rejects unless offer ~1/3.
– Some societies are different, e.g., where giving a gift means power.
– What would you offer/accept? Try it. (Played anonymously. Write offer.)
• Try it table against table. Each table prepares an offer.
- Version 1. The winning table is the one with the greatest total.
- Version 2. A table survives if it winds up with at least $50.
9
The Public Goods Game
• Contributions to a common pot grow—via emergence. The result is
divided among everyone, even free-riders.
• Free riders do better than cooperators/contributors.
• But then cooperation (and public goods) will vanish.
• Punishment is important in sustaining cooperation.
• But how can punishment emerge if it is costly?
Categories of players
•
•
•
•
Loners do not participate; they neither contribute nor benefit.
Defectors do not contribute but benefit.
Cooperators contribute and benefit but do not punish.
Punishers are contributors who also (pay to) punish defectors and simple
cooperators—to prevent simple cooperators from free-riding on punishers.
Which category dominates depends on modeling assumptions.
Games
of Life
Hannelore Brandt, Christoph Hauert, and Karl Sigmund, “Punishing and abstaining
for public goods,” PNAS, Jan 10, 2006. http://www.pnas.org/cgi/reprint/103/2/495
10
Wise crowds: more than the sum of their parts
Traditional wise crowds
• Teams
• Juries
• Democratic voting
Web wise crowd platforms
• Wikis
• Mailing lists
• Chat rooms
• Prediction markets
•
•
•
•
Condorcet Jury Theorem (18th century) example
Five people (a small crowd).
Each person has a 75% chance of being right.
Probability that the majority will be right: ~90%
With 10 people: ~98%. Simple if you think about it.
(James Surowiecki, The Wisdom of Crowds)
(Scott Page, The Difference)
Wise crowd criteria
•
•
Emergence.
Diverse: different skills and information brought to the table.
Decentralized and with independent participants:
Participant
• No one at the top dictates the crowd's answer.
autonomy.
•
•
Each person free to speak his/her own mind and make own decision.
Distillation mechanism: to extract the essence of the crowd's wisdom.
Second slide ahead
11
A wise crowd as assistant and companion
12
Distillation:
making the crowd’s “wisdom” “actionable”
• Elections, polls, etc. Traditional. Many possible
processes, e.g., transferrable ballots, etc.
– Expression of preferences.
– Many online options (and more options).
• Collaboration: wikis and other collaboration tools
(shared spaces), mailing lists, chat rooms, etc.
– Explicit: Generation of new “work products.”
• Here’s a (long!) list of collaborative work environments.
– Implicit: Google’s page rank, “reputations” (e.g., eBay),
“recommendation engines” (e.g., Amazon)
• Knowledge extraction: prediction markets.
13
Prediction markets
Abstract: Prediction markets are markets for
contracts that yield payments based on the
outcome of an uncertain future event, such as
a presidential election. Using these markets as
forecasting tools could substantially improve decision making in the
private and public sectors.
We argue that U.S. regulators should lower barriers to the creation and
design of prediction markets by creating a safe harbor for certain types of
small stakes markets. We believe our proposed change has the potential
to stimulate innovation in the design and use of prediction markets
throughout the economy, and in the process to provide information that
will benefit the private sector and government alike.
Statement issued by 25 world-famous academics. May 2007.
Including: Kenneth Arrow, Daniel Kahneman, Thomas Schelling,
Robert Shiller, Cass Sunstein.
14
Often Beats Alternatives
• Vs. Public Opinion
– I.E.M. beat presidential election polls 451/596 (Berg et al ‘01)
– Re NFL, beat ave., rank 7 vs. 39 of 1947 (Pennock et al ’04)
• Vs. Public Experts
– Racetrack odds beat weighed track experts (Figlewski ‘79)
• If anything, track odds weigh experts too much!
– OJ futures improve weather forecast (Roll ‘84)
– Stocks beat Challenger panel (Maloney & Mulherin ‘03)
– Gas demand markets beat experts (Spencer ‘04)
– Econ stat markets beat experts 2/3 (Wolfers & Zitzewitz ‘04)
from Robin Hanson
• Vs. Private Experts
– HP market beat official forecast 6/8 (Plott ‘00)
– Eli Lily markets beat official 6/9 (Servan-Schreiber ’05)
– Microsoft project markets beat managers (Proebsting ’05)
15
Market mechanisms
• Intrade uses a continual double (bid and asked) auction. (Like stocks).
– Requires high liquidity or a market maker.
– Aggregates information in price; can buy or sell any time.
• Pari-Mutual. Losing bets distributed to winning betters. (Like horse racing).
– Requires neither liquidity nor a market maker.
– Aggregates information as odds. Can’t trade. Prices don’t vary. No profit in
being right early. Best strategy is to wait until the last minute. But that
reduces the amount of information supplied to the pool.
– kahst.
• Market Scoring Rules (Robin Hanson) and Dynamic Pari-Mutuel Market (David
M. Pennock & Mike Dooley).
– Combines pari-mutuel with CDA.
– Benefit for being right early.
– MSR: Inkling, Qmarkets; DPM: Yahoo! Tech Buzz Game.
• List of markets: MidasOracle.org.
16
Split off from
TradeSports
Prediction markets
Contracts: Intrade (Ireland-based): real money or play
money.
But, there is evidence that prediction markets are not efficient.
Panos Ipeirotis
Slate’s
Election
Market
Page
Other Intrade contracts: Current Events > Google Lunar X Prize
Land a privately funded robotic rover on the Moon that is capable of completing several mission objectives,
including roaming the lunar surface for at least 500 meters and sending video, images and data back to the Earth.
17
Concerns and Myths
from Robin Hanson
• Self-defeating prophecies
• Decision selection bias
• Price manipulation
• Rich more “votes”
• Inform “enemies”
• Share less info
• Combinatorics
• Risk distortion
• Moral hazard
• Alarm public
• Embezzle
• Bubbles
• Crowds don’t always beat experts.
• People will not work for trinkets.
• High accuracy is not assured.
• Bozos
• Lies
18
Exploratory behavior: asymmetric warfare
• It is the nature of complex systems
and evolutionary processes that
conflicts become asymmetric.
• No matter how well armored one is …
• there will always be chinks in the
armor, … and something will
inevitably find those chinks.
• The something that finds those
chinks will by definition be
asymmetric since it attacks the
chinks and not the armor.
19
Exploratory behavior: like water finding a way down hill
From a tutorial on the immune system from the National
Cancer Institute: http://www.cancer.gov/cancertopics/understandingcancer/immunesystem.
Microbes attempting to get into your body must first get past your
skin and mucous membranes, which not only pose a physical
barrier but are rich in scavenger cells and IgA antibodies.
Next, they must elude a series of nonspecific defenses—and
substances that attack all invaders regardless of the epitopes they
carry. These include patrolling phagocytes, granulocytes, NK cells,
and complement.
Infectious agents that get past these nonspecific barriers must
finally confront specific weapons tailored just for them. These
include both antibodies and cytotoxic T cells.
Quite a challenge! We are very well defended. But we still get sick!
Some “invaders” will make it past these defenses. The problem is not
even that some get through, it’s that they exploit their success.
How do they find the open pathways? It’s not
“invaders” vs. “defenders.”
Through (evolutionary) exploratory behavior, if
there is a way, some will inevitably find it.
Innovation is the
(disruptive) invader
not the defender.
Innovative organizations make that inevitability work in their favor.
20
Exploratory behavior: recall evolutionary processes
• How can the human genome, with fewer than 25,000 genes
– fill in all the details of the circulatory and nervous systems?
– produce a brain with trillions of cells and synaptic connections?
• Cell growth followed by die-off produce webbing in duck feet and
bat wings but not in human fingers.
• Military strategy of “probing for weakness.”
• Ant and bee foraging.
• Corporate strategy of seeking (or creating) marketing niches.
The general mechanism is:
• Prolifically generate a wide range of possibilities
• Establish connections to new sources of value in the environment.
Mechanism
generation
Function
explore
Purpose
use result
Bottom up
21
Innovative environments
The Internet
• The inspiration for net-centricity and the GIG
• Goal: to bring the creativity of the internet to the DoD
Other innovative environments
• The scientific and technological research process
• The market economy
• Biological evolution
What do
innovative environments
have in common?
22
Innovative environments
Innovation is always the result of an evolutionary process.
• Randomly generate new variants—by combining and
modifying existing ones.
• Select the good ones.
(Daniel Dennett, Darwin's Dangerous Idea)
Requires mechanisms:
• For creating stable and persistent design representations so
that they can serve as the basis for new possibilities.
• For combining and modifying designs.
• For selecting and establishing better ones.
23
Designs in various environments
Recorded as
Created by
How
instantiated
All
bottom-up
Established
Software
Programmers
who know the
techniques
Self-instantiating
By users
Publications
Scientists who
know the
literature
The publication
is the
instantiation
By peer review
Market
economy
Trade secrets
Product
developers
who know the
tricks
Entrepreneurial
manufacturing
By consumers
Biological
evolution
DNA
Combination
and mutation
Reproduction
Whether it finds
a niche
Implicit designs
Construction,
combination
and mutation
Implementation
of a level of
abstraction
Whether it finds
a niche
Internet
Scientific
knowledge
Entities:
nature’s
memes
24
How does this apply to organizations?
To ensure innovation:
Creation and trial
• Encourage the prolific generation and trial
of new ideas.
Establishing successful variants
• Allow new ideas to flourish or wither
based on how well they do.
Sounds simple doesn’t it?
25
Innovation in various environments
New ideas
aren’t the
problem.
Biological
evolution
Entrepreneur
Bureaucracy
Trying them out
Initial funding
Prospect of
failure
Capitalism in
the small.
Nature always
experiments.
Most are failures,
which means
death. (But no
choice given.)
Little needed
for an Internet
experiment.
Perhaps some
embarrassment,
time, money; not
much more.
Proposals,
competition,
forms, etc.
Approvals
Getting
good ideas
Establishment
established
None.
Bottom-up
resource
allocation
defines
success.
Few.
Entrepreneur
wants rewards.
Bottom-up
resource
allocation.
When 100%
Managers have
Mission Success
Far too
other priorities.
is the group goal
many.
Top-down
who wants a
Save ourselves
resource
failure in his/her by spin-doctoring
allocation.
personnel file?
26