universal laws and architectures: theory and lessons from nets, grids, brains, bugs, planes, docs,...
TRANSCRIPT
Universal laws and architectures:Theory and lessons from
nets, grids, brains, bugs, planes, docs, fire, fashion,
art, turbulence, music, buildings, cities, earthquakes, bodies, running, throwing,
Synesthesia, spacecraft, statistical mechanics
slow
inflexible
fragile
waste
eye vision
Act
delay
Control
l noise
error ET j
N
P
+
noise
error
C
Understand this more deeply?
+ Neuroscience
Mechanics+Gravity +
Light +Control theory
exp ln
expT z p
pz pT
eye vision
Act
delay
Control
l noise
error
Understand this more deeply?
+ Neuroscience
Vision
Motion
Robust vision w/motion• Object motion• Self motion
Vision
Motion
eye vision
Act
delay
Control
noiseerror
Fast
Slow
Flexible Inflexible
Vision
Explain this amazing system.
LayeringFeedback
Experiment• Motion/vision control without blurring• Which is easier and faster?
Robust vision with• Hand motion• Head motion
Fast
Slow
VOR
vision
Why?• Mechanism• Tradeoff
Fast
Slow
Flexible Inflexible
VOR
vision
Vestibular Ocular Reflex (VOR)
Mechanism
Tradeoff
Slow
Flexible
vision
eye vision
Actslow
delay
Fast
Inflexible
Slow
Flexible
vision
eye vision
Actslow
delay
Fast
Inflexible
VOR
fast
Slow
Flexible
eye
Act
Fast
Inflexible
VOR
fast
Vestibular Ocular Reflex (VOR)
Slow
Flexible
eye
Act
Fast
Inflexible
VOR
fast
It works in the dark or with your eyes closed, but
you can’t tell.
Slow
Flexible
vision
eye vision
Actslow
delay
Fast
Inflexible
VOR
fast
vision
eye vision
Actslow
delay
VOR
Slow
Flexible
Fast
Inflexible
Illusion
Highly evolved (hidden)
architecture
LayeringFeedback
eye vision
Act
VOR
Layering
AutomaticUnconscious
PartiallyConscious
eye vision
Actslow
delay
VOR
fastLayeringFeedback
vision
eye vision
Actslow
delay
VOR
Slow
Flexible
Fast
Inflexible
Illusion
fast
Architecturelayered/
distributed
vision
eye vision
Actslow
delay
VOR
Slow
Flexible
Fast
Inflexible
Illusion
fast
Architecturelayered/
distributed
Robust or fine-tuned? BothModular or plastic? Both
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Error
+
Very rough cartoon
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Cerebellum
Error
+
ErrorTune gain
Slightly better
gainAOS
AOS = Accessory Optical system
Cortex
eye vision
Actslow
delay
VOR
This fast “forward”
path
Object motion
Head motion
Cerebellum
Error
+
ErrorTune gain
Slightly better
gainAOS
AOS = Accessory Optical system
Needs “feedback”
tuning
Via AOS and cerebellum (not cortex)
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Cerebellum
+
ErrorTune gain
gain
AOS
AOS = Accessory Optical system
noise
Act
Highly evolved (hidden)
architecture
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Cerebellum
+
ErrorTune gain
gain
AOS
AOS = Accessory Optical system
noise
Act
slowest
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Cerebellum
+
ErrorTune gain
gain
AOS
AOS = Accessory Optical system
noise
Act
Delays everywhere
Distributed control
slowest
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Cerebellum
+
ErrorTune gain
gain
AOS
noise
Act
slowestHeterogeneous
delays everywhere
Slow
Flexible
vision
eye vision
Actslow
delay
Fast
Inflexible
VOR
fast
3D + motion
See Marge Livingstone
Color?
Slow
Flexible
vision
eye vision
Actslow
delay
Fast
Inflexible
VOR
fast
See Marge Livingstone
3D + motion
colorvision
Slowest
Stare at the intersection
Stare at the intersection.
Stare at the intersection.
Slow
Flexible
vision
eye vision
Actslow
delay
Fast
Inflexible
VOR
fast
3D + motion
colorvision
Slowest
Slow
Flexible
vision
Fast
Inflexible
VOR
colorvision
Seeing is dreaming• is simulation• requiring “internal model”
gradient
Biased random walk
Ligand
SignalTransduction
Motion Motor
CheY
+CH 3R
ATP ADPP
~
flagellarmotor
Z
Y
PY
~
PiB
B~P
Pi
CW-CH3
ATP
WA
MCPs
WA
+ATT
-ATT
MCPsSLOW
FAST
ligand binding
motor
Bacterial chemotaxis• Internal model necessary for robust chemotaxis• Reality is 3d, but…• Internal model virtual and 1d
Slow
Flexible
vision
Fast
Inflexible
VOR
colorvision
LayeringDistributedFeedback
Slow
Flexible
Fast
Inflexible
Impossible (law)
Architecture
Architecture (constraints that
deconstrain)
General Special
Universal laws and architectures (Turing)
ideal
Slow
Flexible
Fast
Inflexible
General Special
NP(time)
P(time)
analytic
Npspace=Pspace
Decidable
Computation(on and off-line)
Slow
Flexible
Fast
Inflexible
NPPanalytic
PspaceDecidable
Insight
Research progress
Control, OR
Compute
From toys to “real” systems
Slow
Flexible
Fast
Inflexible
NPPanalytic
PspaceDecidable
Insight
Research progress
Control, OR
Compute
Improved algorithms
Slow
Flexible
vision
Fast
Inflexible
VOR
colorvision
LayeringDistributedFeedback
Cortex
eye vision
Actslow
delay
VOR
This fast “forward”
path
Object motion
Head motion
Cerebellum
Error
+
ErrorTune gain
Slightly better
gainAOS
AOS = Accessory Optical system
Needs “feedback”
tuning
Via AOS and cerebellum (not cortex)
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Cerebellum
+
ErrorTune gain
gain
AOS
AOS = Accessory Optical system
noise
Act
sup ( ) ?T
• Real (NP hard)• Complex = LTI (??)• Operator valued or
Noncomm (P)
Scale with dimension?
Uncertainty everywhere
vision
eye vision
Actslow
delay
VOR
Slow
Flexible
Fast
Inflexible
Illusion
Highly evolved (hidden)
architecture
LayeringDistributedFeedback
eye vision
delay .3s
Act
delay
Control
l
1p
l
noiseerror
1x
2x
3x
Model?
• 1 dimension, 4 states?• Other 2 dimensions?• New issues arise
eye vision
Act
delay
Control
noiseerror
ET j
N
1x
2x
3x
eye vision
Act
delay
Control
noiseerror
ET j
N
1x
2x
3xeasy
easy
hard
eye vision
Act
delay
Control
noiseerror
ET j
N
1x
2x
3xeasy
easy
hard
eye vision
Act
delay
Control
noiseerror
ET j
N
Close one eye?
10
100
Length, m
.1 1.5.22 k10
-110
010
1100
101
too fragile
complex
No tradeoff
expensive
fragile
exp lnexp
T z pp
z pT
Slow
Flexible
vision
eye vision
Act slowdelay
Fast
Inflexible
VOR
fastUniversal
laws?
10
100
Length, m
.1 1.5.22
Slow
Flexible
vision
eye vision
Act slowdelay
Fast
Inflexible
VOR
fast
exp ln
expT z p
pz pT
How do these fit together?
slow
flexible
fast
inflexible
How do these fit together?
“waste”“efficient”
fragile
robust
vision
VORdelay
delaysec/m
.01
1
.1
area
.1 1 10diam(m)
.01 1 100
Retinal ganglion axons?
Other myelinated
Why such extreme diversity in axon size and delay?
.01
1
delaysec/m
Aα
C
.1
Aβ
AγAδ
area
.1 1 10diam(m)
.01 1 100
Axon size and
speed
Fast
Small
Slow
Large
delaysec/m
.01
1
.1
area
.1 1 10diam(m)
.01 1 100
Retinal ganglion axons?
Other myelinated
Why such extreme diversity in axon size and delay?
VORNo
diversity in VOR?
Resolution and bandwidth are cheap
speed costs
axon
Same area= same “cost”
axonaxon
2 x resolution(less noise)
axon
x speed(less delay)
Double the area (and cost)?
speed costs Resolution and bandwidth are cheap
1delay
speed
vision
Act
delay
10
100
Speed 1/
.1 1.5.22
.3s
exp p
Slower
axon axon
axon
2 x resolution(less noise)
axon
x
speed doubly expensive
but necessary
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Cerebellum
+
ErrorTune gain
gain
AOS
AOS = Accessory Optical system
noise
Act
Delays everywhere
Distributed control
slowest
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Cerebellum
+
ErrorTune gain
gain
AOS
noise
Act
slowestHeterogeneous
delays everywhere
expE p N
Fast
Slow
Flexible InflexibleReflex
SenseMotor
Prefrontal
Fast
Learn
Evolve
Apps
OS
HW
vision
Act
delay
axon axon
axon
2 x resolution(less noise)
axon
x
speed doubly expensive
but necessary
slow
flexible
fast
inflexible
How do these fit together?
“waste”“efficient”
fragile
robust
vision
VORdelay
speed costs
slow
flexible
fast
inflexible
In general?
wastefulefficient
fragile
robust
slow
flexible
fast
inflexible
fragile
robust
wasteefficient
fragile
robust
slow
flexible
fast
inflexible
fragile
robust
waste
efficient
fragile
robust
Cyber-physicalCyber only
slow
flexible
fast
inflexible
waste
efficient
fragile
robust
Cyber-physical
accessibleaccountableaccurateadaptableadministrableaffordableauditableautonomyavailablecompatiblecomposable configurablecorrectnesscustomizabledebugabledegradabledeterminabledemonstrable
dependabledeployablediscoverable distributabledurableeffective
evolvableextensiblefail
transparentfastfault-tolerantfidelityflexibleinspectableinstallableIntegrityinterchangeabl
einteroperable learnablemaintainable
manageablemobilemodifiablemodularnomadicoperableorthogonalit
yportableprecisionpredictableproducibleprovablerecoverablerelevantreliablerepeatablereproducibleresilientresponsivereusable
safety scalableseamlessself-sustainableserviceablesupportablesecurablesimplestablestandardssurvivable
tailorabletestabletimelytraceableubiquitousunderstandableupgradableusable
efficient
robust
sustainable
PCA Principal Concept Analysis
flexible
fast
efficient
robust
slow
flexible
fast
inflexible
waste
efficient
slow
flexible
fast
inflexible
waste
efficient
Slow
InflexibleFra
gile
Waste
slow
inflexible
fragile
wasteSame picture
Control, OR
CommunicateCompute
Physics
Shannon
Bode
Turing
Gödel
EinsteinHeisenberg
Carnot
Boltzmann
Theory?Deep, but fragmented, incoherent, incomplete
Nash
Von Neumann
Priorities
• Functionality (behavior, semantics)• Robustness
– Uncertain environment and components– Fast (sense, decide, act)– Flexible (adaptable, evolvable)
• Efficiency– Energy– Other resources (make and maintain)
Control, OR
Compute
• Functionality (behavior, semantics)• Robustness
– Uncertain environment and components
– Fast (sense, decide, act)– Flexible (adaptable, evolvable)
• Efficiency– Energy– Other resources (make and maintain)
Function, delay,
robustness most
important
• Functionality (behavior, semantics)• Robustness
– Uncertain environment and components– Fast (sense, decide, act)– Flexible (adaptable, evolvable)
• Efficiency– Energy– Other resources (make and maintain)
Communicate
Physics
Info theory
• Functionality (behavior, semantics)• Robustness
– Uncertain environment and components– Fast (sense, decide, act)– Flexible (adaptable, evolvable)
• Efficiency– Energy– Other resources (make and maintain)
Communicate
Physics
Info theory
Noise, average, asymptotic
Control, OR
Info theoryCompute
Physics
Function, delay,
robustness most
important
Function, delay,
robustness least
important
Control, OR
Info theoryCompute
Physics
Shannon
Bode
Turing
Einstein
Heisenberg
Carnot
Boltzmann
Function, delay,
robustness most
important
Function, delay,
robustness least
important
Dominates “high impact
science” literature
Slow
Flexible
Fast
Inflexible
NPPanalytic
PspaceDecidable
Insight
Research progress
Control, OR
Compute
From toys to “real” systems
Slow
Flexible
Fast
Inflexible
NPPanalytic
PspaceDecidable
Insight
Research progress
Control, OR
Compute
Improved algorithms
Control, OR
Info theoryCompute
Physics
Shannon
Bode
Turing
Einstein
Heisenberg
Carnot
Boltzmann
Function, delay,
robustness most
important
Function, delay,
robustness least
important
Function, delay,
robustness most
important
Function, delay,
robustness least
important
Control, OR
CommunicateCompute
Physics
Shannon
Bode
Turing
New progress!
Control, OR
Compute
Physics
Bode
Turing
New progress!
Coherent structures and turbulent drag
Function, delay,
robustness least
important
Shear flow turbulence• Exhaustively studied
– Extensive experiments and big data– Detailed models and big simulations– Great! But all just deepen the mystery
• Perfectly illustrates robust/fragile• Without which? Bewilderment.
Physics of Fluids (2011)
Physics of Fluids (2011)
wU
z x
y
uzx
yFlow
upflowhigh-speed
region
downflowlow speed
streak
Blunted turbulent velocity profile
Laminar
Turbulent
wU3D coupling
Coherent structures and turbulent drag
wasteful
fragile Laminar
Turbulent
efficient
robust
Laminar
Turbulent
wU
?
Control?
Fundamentals!
Control, OR
Compute
Bode
Turing
New progress!
Communications for control,
computing, biology
Function, delay,
robustness least
important
Internal delays
SenseMotor
Prefrontal
Fast
Learn
Reflex
Evolve
vision
VOR
Communications for control,
computing, biology
Internal delays
New theory: Distributed control with delays everywhere.
Sense/Act
Compute/communicate
Physical
Info Thry
Control, OR
Compute
PhysicsOrthophysics (Eng/Bio/Math)
Optimization Statistics
Comms for Comp/Cntrl/Bio
Theory
wasteful
fragile Laminar
Turbulent
efficient
robust ?
Control?
Info Thry
PhysicsOrthophysics (Eng/Bio/Math)
Laminar
Turbulent
uzx
yFlow
Doyle, Csete, Proc Nat Acad Sci USA, JULY 25 2011
For more infoMost accessible
No mathReferences
wasteful
fragile
efficient
robust
Impossible (law)
Architecture
Universal laws and architectures
Flexibly achieves what’s possible
Efficiency/instability/layers/feedback
• Sustainable infrastructure? (e.g. smartgrids)• Money/finance/lobbyists/etc• Industrialization• Society/agriculture/weapons/etc• Bipedalism• Maternal care• Warm blood• Flight• Mitochondria• Oxygen• Translation (ribosomes)• Glycolysis (2011 Science)
Major transitions
Evolvability?
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
SensoryMotor
Prefrontal
Striatum
Reflex
LearningSoftware
Hardware
Digital
Analog
Horizontal Gene
Transfer
Horizontal App
Transfer
Horizontal Meme
Transfer
Accelerating evolution
Requires shared “OS”
Horizontal Gene
Transfer
Sequence ~100 E Coli (not chosen randomly)• ~ 4K genes per cell• ~20K different genes in total (pangenome)• ~ 1K universally shared genes • ~ 300 essential (minimal) genes
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
SensoryMotor
Prefrontal
Striatum
Reflex
LearningSoftware
Hardware
Digital
Analog
Horizontal Gene
Transfer
Horizontal App
Transfer
Horizontal Meme
Transfer
What can go wrong?
Horizontal Bad Gene Transfer
Horizontal Bad App Transfer
Horizontal Bad Meme
Transfer
Parasites &
Hijacking
Fragility?
Exploiting layered
architecture
Virus
Virus
Meme
Horizontal Bad Meme
Transfer
Our greatest fragility?
Meme
Many human beliefs are:• False• Unhealthy
Slow
Flexible
vision
Fast
Inflexible
VOR
Slow
Flexible
vision
eye vision
Actslow
delay
Fast
Inflexible
VOR
fast
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Cerebellum
+
ErrorTune gain
gain
AOS
AOS = Accessory Optical system
noise
Act
Delays everywhere
Distributed control
slowest
Slow
Flexible
vision
eye vision
Actslow
delay
Fast
Inflexible
VOR
fast
Vision
-slow
VOR
fast
Fast
Slow
Flexible Inflexible
slower
Mixed
Redraw(not anatomical)
Vision
-3d+motionB&W
slowVOR
fast
Fast
Slow
Flexible Inflexible
slower
Objects
Mixed
Redraw(not anatomical)
3d+motionB&W
slow
Fast
Slow
Flexible Inflexible
slower
Objects
-3d+motionB&W
slowVOR
fast
Fast
Slow
Flexible Inflexible
Faces
slower
Objects
Mixed
Not sure how to draw this…
Chess experts• can reconstruct entire chessboard with < ~ 5s inspection• can recognize 1e5 distinct patterns• can play multiple games blindfolded and simultaneous• are no better on random boards
(Simon and Gilmartin, de Groot)
www.psywww.com/intropsych/ch07_cognition/expertise_and_domain_specific_knowledge.html
Specific brain lesions can cause “blindness” to• Faces• Fruit or Animals or Tools or ??• Left side of body• Movement• Nonmovement• ???
Fast
Slow
Flexible Inflexible
Faces
slower
Objects
Not sure how to draw this…
Learning+evolution
When needed, even wasps can do it.
• Polistes fuscatus can differentiate among normal wasp face images more rapidly and accurately than nonface images or manipulated faces. • Polistes metricus is a close relative lacking facial recognition and specialized face learning. • Similar specializations for face learning are found in primates and other mammals, although P. fuscatus represents an independent evolution of specialization. • Convergence toward face specialization in distant taxa as well as divergence among closely related taxa with different recognition behavior suggests that specialized cognition is surprisingly labile and may be adaptively shaped by species-specific selective pressures such as face recognition.
Fig. 1 Images used for training wasps.
M J Sheehan, E A Tibbetts Science 2011;334:1272-1275Published by AAAS
vision
eye vision
Actslow
delay
VOR
Slow
Flexible
Fast
Inflexible
Illusion
Highly evolved (hidden)
architecture
Seeing is dreaming
-3d+motionB&W
slowVOR
fast
Fast
Slow
Flexible Inflexible
Faces
slower
Objects
Mixed
Not sure how to draw this…
Learning+evolution
Sense
Universal tradeoffs?
Fast
Slow
Flexible Inflexible
Motor
Prefrontal
Fast
Learn
Reflex
Evolve
Learning can be very slow.
Evolution is even slower.
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
SensoryMotor
Prefrontal
Striatum
Reflex
LearningSoftware
Hardware
Digital
Analog
Horizontal Gene
Transfer
Horizontal App
Transfer
Horizontal Meme
Transfer
Accelerating evolution
Requires shared “OS”
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
SensoryMotor
Prefrontal
Striatum
Reflex
LearningSoftware
Hardware
Digital
Analog
Gene Transfer
App Transfer
Meme Transfer
Some genes, apps, ideas
are “invented”
But most bacteria and
people get most of them from
others
costly
fragile
efficient
robust
Learning
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Cerebellum
+
ErrorTune gain
gain
AOS
AOS = Accessory Optical system
noise
Act
Cortex
eye vision
Actslow
delay
VOR
fast
Object motion
Head motion
Cerebellum
+
Error
Tune gain
gain
AOS
AOS = Accessory Optical system
noise
Act
Reflex
SensoryMotor
Prefrontal
Striatum
SlowFlexible
Learning
Ashby & Crossley
Slow
Reflex(Fastest,
LeastFlexible)
Extreme heterogeneity
SensoryMotor
Prefrontal
Striatum
Fast
FastInflexible
SlowFlexible
Learning
Reflex(Fastest,
LeastFlexible)
Ashby & Crossley
Learning can be very slow.
Sense
Fast
Slow
Flexible Inflexible
Motor
Prefrontal
Fast
Learn
Reflex
Evolve
HMTTeach
Horizontal Meme
Transfer
Sense
Fast
Slow
Flexible Inflexible
Motor
Prefrontal
Fast
Learn
Reflex
Evolve
Horizontal Hardware Transfer
HHT
Sense
Fast
Slow
Flexible Inflexible
Motor
Prefrontal
Fast
Learn
Reflex
Evolve
Apps
OS
HW
HMT
HHT
Sense
Fast
Slow
Flexible Inflexible
Motor
Prefrontal
Fast
Learn
Reflex
Evolve
HMT
Soma
Reactive
Adaptive
Contextual
Sense
Fast
Slow
Flexible Inflexible
Motor
Prefrontal
Fast
Learn
Reflex
Evolve
Apps
OS
HW
FastCostly
SlowCheap
Flexible Inflexible
General Special
Layered Architecture
AppsOS
HW
Digital
Lumped
Distrib.
OSHW
Digital
Lumped
Distrib.
DigitalLumped
Distrib.
LumpedDistrib. Distrib.
Any layer needs all lower layers.
Fast
Slow
Flexible Inflexible
Apps
OS
HW
AppsOS
HardwareDigital
LumpedDistributed
Tech implications/extensions
Slow
Flexible
Fast
Inflexible
Impossible (law)
Architecture
Architecture (constraints that
deconstrain)
General Special
Universal laws and architectures (Turing)
ideal
Slow
Flexible
Fast
Inflexible
General Special
NP(time)
P(time)
analytic
Npspace=Pspace
Decidable
Computation(on and off-line)
Slow
Flexible
Fast
Inflexible
General Special
NP
P
analytic
Pspace
Decidable
Computation(on and off-line)
InsightAccessibleVerifiable
Slow
Flexible
Fast
Inflexible
NPPanalytic
PspaceDecidable
Insight
Research progress
Control, OR
Compute
From toys to “real” systems
Slow
Flexible
Fast
Inflexible
NPPanalytic
PspaceDecidable
Insight
Research progress
Control, OR
Compute
Improved algorithms
P(time)(delay)
Pspace(memory, bandwidth)
Resource cost vs effectiveness
Cheap Costly
Strong
Weak
To make
In use
Talk
Energy
Pspace(memory, bandwidth)
Cheap
Strong
To make
In use
1 TB < $100
$50
$??
1 TB = 1e12B
P(time)
Pspace(memory, bandwidth)
What dominates tradeoffs
Cheap Costly
Strong
Weak
To make
In use
Talk
Energy
time
What dominates tradeoffs
CostlyCheap
Fast
Slow
Energy
time
What dominates tradeoffs
CostlyCheap
Fast
Slow
Energy
Flexible
Inflexible
time
What dominates tradeoffs
Fast
Slow
Flexible
Inflexible
Slow
Flexible
Fast
Inflexible
General Special
NP(time)
P(time)
analytic
Pspace
Decidable
Computation(on and off-line)
Slow
Flexible
Fast
Inflexible
General Special
NP(time)
P(time)
analytic
Pspace
Decidable
Universal architecture
Universal TM
10
100
Length, m
.1 1.5.22 k10
-110
010
1100
101
too fragile
complex
No tradeoff
expensive
fragile
exp lnexp
T z pp
z pT
Slow
Flexible
vision
eye vision
Act slowdelay
Fast
Inflexible
VOR
fastSlow
Flexible
Fast
Inflexible
General Special
NPPanalytic
Pspace
Decidable
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
SensoryMotor
Prefrontal
Striatum
Reflex
LearningSoftware
Hardware
Digital
Analog
Horizontal Gene
Transfer
Horizontal App
Transfer
Horizontal Meme
Transfer
Accelerating evolution
Requires shared “OS”
Sense
Fast
Slow
Flexible Inflexible
Motor
Prefrontal
Fast
Learn
Reflex
Evolve
Apps
OS
HW
HMT
Fast
Slow
Flexible Inflexible
General Special
Apps
OS
HW
Exploiting layered
architecture
Core protocols
• OS seems to confuse• Replace with “core protocols”?• Examples different than OS or translation:
– Core metabolism– 2 component signal transduction
• Highly conserved/constrained core (knot)• Highly diverse/deconstrained edges
153
slow
flexible
fast
inflexible
waste
efficient
fragile
robust
slow
flexible
fast
inflexible
waste
efficient
body
brain
Combo
Sense
Fast
Slow
Flexible Inflexible
Motor
Prefrontal
Fast
Learn
Reflex
Evolve
HMT
Soma
Reactive
Adaptive
Contextual
Combo
Slow
Fast
Flexible InflexibleGeneral Special
AppsOSHWDig.Lump.Distrib.
OSHWDig.Lump.Distrib.
DigitalLump.Distrib.
LumpedDistrib. Distrib.
HGT DNA repair Mutation DNA replication Transcription
Translation
Metabolism Signal
SenseMotor
Prefrontal
Fast
Learn
Reflex
Evolve
vision
VOR
Slow
Fast
Flexible InflexibleGeneral Special
AppsOSHWDig.Lump.Distrib.
OSHWDig.Lump.Distrib.
DigitalLump.Distrib.
LumpedDistrib. Distrib.
Apps
OS
HW
Fast
Slow
Flexible Inflexible
General Special
Apps
OS
HW
Horizontal App
Transfer
Horizontal Hardware Transfer
Fast
Slow
Flexible Inflexible
General Special
Apps
OS
HW
OS
Horizontal App
Transfer
Horizontal Hardware Transfer
Constrained
Deconstrained
Deconstrained
Horizontal Transfer
Horizontal Transfer
Constrained
Deconstrained
Deconstrained
Con
stra
ined
Dec
onst
rain
ed
Dec
onst
rain
ed
food
Glucose
Oxygen
OrgansTissuesCellsMolecules
Universal metabolic system
Blood
EfficientRobust
Evolvable
food
Glucose
Oxygen
OrgansTissuesCellsMolecules
Blood
EfficientRobust
Evolvable
Constrained
DeconstrainedDeconstrained
EfficientRobust
Evolvable
Constrained
DeconstrainedDeconstrained
DiverseDiverse
EfficientRobust
Evolvable
Constrained
DeconstrainedDeconstrained
DiverseDiverse
OS
Deconstrained(Hardware)
Deconstrained(Applications)
Constrained
universal architectures
Fast
Slow
Flexible Inflexible
General Special
Apps
OS
HW
Horizontal App
Transfer
Horizontal Hardware Transfer
FastCostly
SlowCheap
Flexible Inflexible
HGT DNA repair
Mutation DNA replication
Transcription Translation
MetabolismSignal…
Layered Architecture
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
FastCostly
SlowCheap
Flexible Inflexible
HGT DNA repair
Mutation DNA replication
Transcription Translation
MetabolismSignal…
Layered Architecture RNA
mRNA
RNApTranscription
Gene
AminoAcids
Proteins
RibosomesTranslation
Metabolism Products
Core Protocols
FastCostly
SlowCheap
Flexible
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
Metabolism
FastCostly
SlowCheap
Flexible
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
Metabolism
FastCostly
SlowCheap
Flexible
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
Signal Transduction
FastCostly
SlowCheap
Flexible
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
Signal Transduction
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
Tra
nsm
itte
r
Rec
eive
r
Variety ofLigands &Receptors
Variety ofresponses
• 50 such “two component” systems in E. Coli• All use the same protocol
- Histidine autokinase transmitter- Aspartyl phospho-acceptor receiver
• Huge variety of receptors and responses• Also multistage (phosphorelay) versions
Signal transduction
Tran
smit
ter
Rec
eive
r
Ligands &Receptors
Responses
Shared protocols
Flow of “signal”
Recognition,specificity
• “Name resolution” within signal transduction• Transmitter must locate “cognate” receiver and avoid non-cognate receivers• Global search by rapid, local diffusion• Limited to very small volumes
Tran
smit
ter
Rec
eive
r
Ligands &Receptors
Responses
Tran
smit
ter
Rec
eive
r
Ligands &Receptors
Responses
Tran
smit
ter
Rec
eive
r
Ligands &Receptors
Responses
Shared protocols
Flow of “signal”
Recognition,specificity
Variety ofLigands &Receptors
Variety ofresponses
Variety ofLigands &Receptors
Variety ofLigands &Receptors
Variety ofLigands &Receptors
Variety ofLigands &Receptors
Variety ofLigands &Receptors
Variety ofresponses
Variety ofLigands &Receptors
Variety ofresponses
Variety ofLigands &Receptors
Variety ofresponses
Variety ofLigands &Receptors
Variety ofresponses
Variety ofLigands &Receptors
Variety ofresponses
Variety ofLigands &Receptors
Variety ofresponses
Variety ofLigands &Receptors
Variety ofresponses
Variety ofLigands &Receptors
Variety ofresponses
Huge
variety
Huge variety
Tran
smit
ter
Rec
eive
rTran
smit
ter
Rec
eive
rTran
smit
ter
Rec
eive
r
Recognition,specificity
Huge variety• Combinatorial• Almost digital• Easily reprogrammed• Located by diffusion
Tran
smit
ter
Rec
eive
rTr
ansm
itte
r
Rec
eive
rTr
ansm
itte
r
Rec
eive
r
Flow of “signal”
Limited variety• Fast, analog (via #)• Hard to change
Reusable in different pathways
InternetsitesUsers
Tran
smit
ter
Rec
eive
r
Ligands &Receptors
Responses
Shared protocols
Flow of “signal”
Recognition,specificity
Note: Any wireless system and the Internet to which it is connected work the same way.
Lap
top
Ro
ute
r
Flow of packets
Recognition,Specificity (MAC)
FastCostly
SlowCheap
Flexible Inflexible
HGT DNA repair
Mutation DNA replication
Transcription Translation
Metabolism Signaling
Layered Architecture
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
gradient
Biased random walk
Ligand
SignalTransduction
Motion Motor
CheY
+CH 3R
ATP ADPP
~
flagellarmotor
Z
Y
PY
~
PiB
B~P
Pi
CW-CH3
ATP
WA
MCPs
WA
+ATT
-ATT
MCPsSLOW
FAST
ligand binding
motor
More necessity and robustness
• Integral feedback and signal transduction (bacterial chemotaxis, G protein) (Yi, Huang, Simon)
• Example of “exploratory process”
Random walk
Ligand Motion Motor
Bacterial chemotaxis
gradient
Biased random walk
Ligand
SignalTransduction
Motion Motor
CheY
Ligand
SignalTransduction
Motion Motor
CheY
Ligand
SignalTransduction
Motion Motor
CheY
Component of feedback controller
Ligand Motion Motor
CheY
Tra
nsm
itte
r
Rec
eive
r
Receptor Response
Component of feedback controller
PMF
++
++
From Taylor, Zhulin, Johnson
Variety of receptors/
ligands Common cytoplasmic
domains
Common signal carrier
Common energy carrier
Tra
nsm
itte
r R
ecei
ver
MotorVarietyof receptors& ligands
Ligand Motion Motor
CheY
Tra
nsm
itte
r R
ecei
ver
MotorVarietyof receptors& ligands
CheY
Tra
nsm
itte
r
Rec
eive
r
Receptor Response
Integral feedback internal to signaling
network
Ligand Motion Motor
CheY
Tra
nsm
itte
r
Rec
eive
r
Receptor Response
Integral feedback internal to signaling
network
Main feedback control loop
Ligand
SignalTransduction
Motion Motor
CheY
gradient
Biased random walk
Ligand
SignalTransduction
Motion Motor
CheY
Tumbling bias
pCheY
SignalTransduction
Motor
Perfect adaptation is necessary …
ligandpCheY
Tumbling bias
ligand
Perfect adaptation is necessary …
…to keep CheYp in the responsive range of the motor.
pCheY pCheY
Tumbling bias
Ligand
F
y+
d
F ln(S)
1
1
yS
d F
ˆ ( ) ˆ( ) , (0) 0
(0)
(0) 0
F sF s F
s
F
S
Integral feedback
Tran
smit
ter
Rec
eive
r
Variety ofLigands &Receptors
Response
DiverseDeconstrained
(Hardware)
Deconstrained(Applications)
Diverse
Layered architectures
ConstrainedBut hidden
Apps
OS
HW
Core Protocols
?
Deconstrained(Hardware)
Deconstrained(Applications)
Next layered architectures (SDN)
Constrained Optimize & control, share, virtualize, manage resources
CommsMemory, storageLatencyProcessingCyber-physical
Few global variables
Don’t cross layers
The essence of layering
Physical resources
Optimize & control, share, virtualize, manage resources
Optimize & control, share, virtualize, manage resources
Optimize & control, share, virtualize, manage resources
Optimize & control, share, virtualize, manage resources
Applications
The essence of networking
Physical
Control
Control
Apps
Control
Physical
Apps
Control
Physical
Apps
Virtual
Prosen-cephalon
Telen-cephalon
Rhinencephalon, Amygdala,
Hippocampus, Neocortex, Basal ganglia, Lateral
ventricles
Dien-cephalon
Epithalamus, Thalamus,
Hypothalamus, Subthalamus, Pituitary
gland, Pineal gland, Third ventricle
Brain stem
Mesen-cephalon
Tectum, Cerebral peduncle, Pretectum, Mesencephalic duct
Rhomb-encephalon
Meten-cephalon
Pons, Cerebellum
Myelen-cephalon
Medulla oblongata
Spinal cord
CNS HW “stack”
?
Diverse?Deconstrained
(Hardware)
Deconstrained?(Applications)
Diverse
Layered architectures??
ConstrainedBrain
Fast
Slow
Flexible Inflexible
General Special
Apps
OS
HW
Horizontal App
Transfer
Horizontal Hardware Transfer
FastCostly
SlowCheap
Flexible Inflexible
General Special
Layered Architecture
AppsOSHW
DigitalLumpedDistrib.
OSHW
DigitalLumpedDistrib.
DigitalLumpedDistrib.
LumpedDistrib. Distrib.
Any layer needs all lower layers.
0. HGT1. DNA repair2. Mutation 3. DNA replication4. Transcription5. Translation6. Metabolism7. Signal transduction8. …
AppsOS
HardwareDigital
LumpedDistributed
Any layer needs all lower layers.
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
Horizontal Gene
Transfer
Sequence ~100 E Coli (not chosen randomly)• ~ 4K genes per cell• ~20K different genes in total (pangenome)• ~ 1K universally shared genes • ~ 300 essential (minimal) genes
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
FastCostly
SlowCheap
Flexible Inflexible
General Special
HGT DNA repair
Mutation DNA replication
Transcription Translation
MetabolismSignal…
Layered Architecture
Slow
Fast
Flexible InflexibleGeneral Special
AppsOSHWDig.Lump.Distrib.
OSHWDig.Lump.Distrib.
DigitalLump.Distrib.
LumpedDistrib. Distrib.
HGT DNA repair Mutation DNA replication Transcription
Translation Metabolism
Signal
SenseMotor
Prefrontal
Fast
Learn
Reflex
Evolve
vision
VOR
Slow
Fast
Flexible InflexibleGeneral Special
AppsOSHWDig.Lump.Distrib.
OSHWDig.Lump.Distrib.
DigitalLump.Distrib.
LumpedDistrib. Distrib.
HGT DNA repair Mutation DNA replication Transcription
Translation
Metabolism Signal
SenseMotor
Prefrontal
Fast
Learn
Reflex
EvolveVOR
Many systems/organisms only use lower layers
LumpedDistrib. Distrib.
Metabolism
Signal
SenseMotor Fast
Reflex
EvolveVOR
Many systems/organisms only use lower layers
Analog
Red blood cellsNormocephalic anucleosis
Most metazoans
SenseMotor Fast
Reflex
Evolve
Analog?
Most metazoans
LumpedDistrib. Distrib.
Metabolism
Signal
SenseMotor Fast
Reflex
Evolve
Many systems/organisms only use lower layers
Analog
Red blood cellsNormocephalic anucleosis
Most metazoans
200 microns
Denucleated neurons
Megaphragma mymaripenne
5 day lifespan7,400 neurons vshousefly 340,000 honeybee 850,000.
Normocephalic anucleosis
Slow
Fast
Flexible InflexibleGeneral Special
AppsOSHWDig.Lump.Distrib.
OSHWDig.Lump.Distrib.
DigitalLump.Distrib.
LumpedDistrib. Distrib.
HGT DNA repair Mutation DNA replication Transcription
Translation Metabolism
Signal
SenseMotor
Prefrontal
Fast
Learn
Reflex
Evolve
vision
VOR
10
100
Length, m
.1 1.5.22 k10
-110
010
1100
101
too fragile
complex
No tradeoff
expensive
fragile
exp lnexp
T z pp
z pT
Slow
Flexible
vision
eye vision
Act slowdelay
Fast
Inflexible
VOR
fastSlow
Flexible
Fast
InflexibleGeneral Special
NPPanalytic
PspaceDecidable
control feedback
RNA
mRNA
RNApTranscription
Other Control
Gene
AminoAcids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transductionATP
DNA
New gene
SensoryMotor
Prefrontal
Striatum
Reflex
LearningSoftware
Hardware
Digital
Analog
Horizontal Gene
Transfer
Horizontal App
Transfer
Horizontal Meme
Transfer
What can go wrong?
Before exploring
mechanisms
Horizontal Bad Gene Transfer
Horizontal Bad App Transfer
Horizontal Bad Meme
Transfer
Parasites &
Hijacking
Fragility?
Exploiting layered
architecture
Virus
Virus
Meme
Horizontal Bad Meme
Transfer
Our greatest fragility?
Meme
Many human beliefs are:• False• Unhealthy
Horizontal Bad Meme
Transfer
Our greatest fragility?
Meme
Many human beliefs are:• False• Unhealthy
• Masculinity• Compassion
Horizontal Transfer
Horizontal Transfer
Constrained
Deconstrained
Deconstrained
Universal architectures
Universal laws
Fast
Slow
Flexible InflexibleGeneral Special
SW Apps
OS
HW