Page 1
John Doyle 道陽
Jean-Lou Chameau Professor
Control and Dynamical Systems, EE, & BioE
tech 1 # Ca
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
Page 2
• Nets
• Grids (cyberphys)
• Brains
• Bugs (microbes)
• Medical physiology
• Lots of aerospace
• Wildfire ecology
• Earthquakes
• Physics:
– turbulence,
– stat mech (QM?)
• “Toy”:
– Lego
– clothing, fashion
• Buildings, cities
• Ants • Synesthesia
A rant
Case StudyI
Page 3
App App
IPC
Global
and direct
access to
physical
address!
DNS
IP addresses
interfaces
(not nodes)
caltech.edu?
131.215.9.49
Page 4
App App
IPC
Global
and direct
access to
physical
address!
Robust?
• Secure
• Scalable
• Verifiable
• Evolvable
• Maintainable
• Designable
• …
DNS
IP addresses
interfaces
(not nodes)
Page 5
accessible accountable accurate adaptable administrable affordable auditable autonomy available credible process
capable compatible composable configurable correctness customizable debugable degradable determinable demonstrable
dependable deployable discoverable distributable durable effective efficient evolvable extensible fail transparent fast fault-tolerant fidelity flexible inspectable installable Integrity interchangeable interoperable learnable maintainable
manageable mobile modifiable modular nomadic operable orthogonality portable precision predictable producible provable recoverable relevant reliable repeatable reproducible resilient responsive reusable robust
safety scalable seamless self-sustainable serviceable supportable securable simplicity stable standards
compliant survivable sustainable tailorable testable timely traceable ubiquitous understandable upgradable usable
Requirements on systems and architectures
Page 6
accessible accountable accurate adaptable administrable affordable auditable autonomy available compatible composable configurable correctness customizable debugable degradable determinable demonstrable
dependable deployable discoverable distributable durable effective
evolvable extensible fail transparent fast fault-tolerant fidelity flexible inspectable installable Integrity interchangeable interoperable learnable maintainable
manageable mobile modifiable modular nomadic operable orthogonality portable precision predictable producible provable recoverable relevant reliable repeatable reproducible resilient responsive reusable
safety scalable seamless self-sustainable serviceable supportable securable simple stable standards survivable
tailorable testable timely traceable ubiquitous understandable upgradable usable
efficient
robust
sustainable
Sustainable robust + efficient
Page 7
accessible accountable accurate adaptable administrable affordable auditable autonomy available compatible composable configurable correctness customizable debugable degradable determinable demonstrable
dependable deployable discoverable distributable durable effective
evolvable extensible fail
transparent fast fault-tolerant fidelity flexible inspectable installable Integrity interchangeabl
e interoperable learnable maintainable
manageable mobile modifiable modular nomadic operable orthogonalit
y portable precision predictable producible provable recoverable relevant reliable repeatable reproducible resilient responsive reusable
safety scalable seamless self-sustainable serviceable supportable securable simple stable standards survivable
tailorable testable timely traceable ubiquitous understandable upgradable usable
efficient
robust
sustainable
Simple
dichotomous
tradeoff pairs
PCA Principal Concept Analysis
wasteful
fragile
efficient
robust
Page 8
wasteful
fragile
efficient
robust
Layering
Feedback
Page 9
Layers/feedback in biology
Working backwards
• Money/finance/lobbyists/etc
• Society/agriculture/weapons/etc
• Bipedalism
• Maternal care
• Warm blood
• Flight
• Mitochondria
• Translation (ribosomes)
• Glycolysis (see 2011 Science paper)
Page 10
Robust
Fragile
Universal
laws
wasteful
efficient
Page 11
Slow
Flexible
Fast
Inflexible
Robust
Fragile
Universal
laws
wasteful
efficient
Page 12
Slow
Flexible
Fast
Inflexible
Robust
Fragile
Universal
laws
Page 13
Slow
Flexible
Fast
Inflexible
Ideal
Fragile Architecture?
Page 14
Slow
Flexible
Fast
Inflexible
Ideal
Fragile
Architecture (constraints that
deconstrain)
Page 15
Slow
Flexible
Fast
Inflexible
Architecture
(constraints that
deconstrain)
General Special
Universal laws and architectures
(Turing)
ideal
Page 16
Fast
Slow
Flexible Inflexible
Apps
OS
HW
Apps
OS
Hardware
Digital
Lumped
Distributed
Tech implications/extensions
Page 17
Layered
Architecture
Apps
OS
Hardware
Digital
Lumped
Distributed
Any layer
needs all
lower layers.
Page 18
Layered architectures 101
Apps
OS
HW
Operating System
Apps
OS
Hardware
Digital
Lumped
Distributed
Page 19
OS
Diverse
Deconstrained
(Hardware)
Deconstrained
(Applications)
Diverse Horizontal
App
Transfer
Horizontal
HW
Transfer
Layered architectures
Apps
OS
HW
Page 20
OS
Layered architectures
Minimal
diversity
and
change
Apps
OS
HW
Page 21
Diverse
Deconstrained
Deconstrained
Diverse
Horizontal
Transfer
Horizontal
Transfer
diversity and change
diversity and change
Page 22
Diverse
Deconstrained
(Hardware)
Deconstrained
(Applications)
Diverse Horizontal
App
Transfer
Horizontal
HW
Transfer
Layered architectures
Apps
OS
HW
Maximal
diversity
and
change
Page 23
Diverse
Deconstrained
(Hardware)
Deconstrained
(Applications)
Diverse
Layered architectures
Constrained
But hidden
Apps
OS
HW
Core
Protocols
Page 24
Fast
Slow
Flexible Inflexible
General Special
Apps
OS
HW
Page 25
Fast
Slow
Flexible Inflexible
General Special
Apps
OS
HW
Page 26
Digital
Lumped Lumped.
Apps
OS
Hardware
Digital
Lumped
Act
Decide
Sense
Page 27
Act
Sense
Apps
OS
Hardware
Digital
Lumped
Decide
Slow
Page 28
Fast
Costly
Slow
Cheap
Flexible Inflexible
General Special
Layered
Architecture
Apps
OS
HW
Digital
Lumped
Distrib.
OS
HW
Digital
Lumped
Distrib.
Digital
Lumped
Distrib.
Lumped
Distrib.
Distrib.
Any layer needs
all lower layers.
Page 29
?
Diverse
Deconstrained
(Hardware)
Deconstrained
(Applications)
Diverse
Layered architectures
Constrained
Apps
OS
HW
Minimal
diversity
Layered
architectures
make robustness
and evolvability
compatible
Ask me about
hourglasses
and bowties
Page 30
App App
IPC
Global
and direct
access to
physical
address!
Robust?
• Secure
• Scalable
• Verifiable
• Evolvable
• Maintainable
• Designable
• …
DNS
IP addresses
interfaces
(not nodes)
Page 31
Physical
IP
TCP
Application
“Issues” (hacks)
• VPNs
• NATS
• Firewalls
• Multihoming
• Mobility
• Routing table size
• Overlays
• …
Page 32
Original design challenge?
TCP/
IP
Deconstrained
(Hardware)
Deconstrained
(Applications)
Constrained • Expensive mainframes
• Trusted end systems
• Homogeneous
• Sender centric
• Unreliable comms
Facilitated wild evolution
Created
• whole new ecosystem
• completely opposite
Networked OS
Page 33
Original design challenge?
TCP/
IP
Deconstrained
(Hardware)
Deconstrained
(Applications)
Constrained Incomplete
• Layering (IP)
• Feedback (TCP)
• No theory
Networked OS
Page 34
Chiang, Low, Calderbank, and Doyle
Page 35
wasteful
fragile
efficient
robust
Layering
Feedback
Page 36
?
Deconstrained
(Hardware)
Deconstrained
(Applications)
Next layered arches (SDN? OpenDaylight?)
Constrained Optimize: control, share,
virtualize, manage resources
Comms
Memory, storage
Latency
Processing
Cyber-physical
Few global variables
Don’t cross layers
Page 37
?
Deconstrained
(Hardware)
Deconstrained
(Applications)
Next layered architectures
Constrained Optimize: control, share,
virtualize, manage resources
Comms
Memory, storage
Latency
Processing
Cyber-physical
Page 38
Fast
Slow
Flexible Inflexible
Vision
Explain this
amazing
system.
Layering
Feedback
Page 39
Robust vision with motion
Vision
Motion
• Neuroscience motivation
Page 40
Experiment • Motion/vision control without blurring
• Which is easier?
Page 41
Fast
Slow
VOR
vision
Why?
• Mechanism
• Tradeoff
Page 42
Fast
Slow
Flexible Inflexible
VOR
vision
Vestibular
Ocular
Reflex
(VOR)
Mechanism
Tradeoff
Page 43
Slow
Flexible
vision
eye vision
Act slow
delay
Fast
Inflexible
Page 44
Slow
Flexible
vision
eye vision
Act slow
delay
Fast
Inflexible
VOR
fast
Page 45
Slow
Flexible
eye
Act
Fast
Inflexible
VOR
fast
Vestibular
Ocular
Reflex
(VOR)
Page 46
Slow
Flexible
eye
Act
Fast
Inflexible
VOR
fast
Page 47
Slow
Flexible
vision
eye vision
Act slow
delay
Fast
Inflexible
VOR
fast
Page 48
vision
eye vision
Act slow
delay
VOR
Slow
Flexible
Fast
Inflexible
Illusion
Highly
evolved
(hidden)
architecture
Layering
Feedback
Page 49
eye vision
Act
VOR
Layering
Automatic
Unconscious
Partially
Conscious
Page 50
eye vision
Act slow
delay
VOR
fast
Layering
Feedback
Page 51
vision
eye vision
Act slow
delay
VOR
Slow
Flexible
Fast
Inflexible
Illusion
fast Architecture
Page 52
Layers/feedback in biology
Working backwards
• Money/finance/lobbyists/etc
• Society/agriculture/weapons/etc
• Bipedalism
• Maternal care
• Warm blood
• Flight
• Mitochondria
• Translation (ribosomes)
• Glycolysis (see 2011 Science paper)
Page 53
Universal laws and architectures: brains, bugs, networks, physiology,
grids, medicine, wildfire, turbulence, literature, fashion, dance, earthquakes,
art, music, Lego, buildings, cities vision
Act
delay + Neuroscience
Balancing
an inverted
pendulum
Mechanics+
Gravity +
Light +
2 2
0
1ln
ln
pT j d
p
z pp
z p
Page 54
easy
Law #1 : Mechanics
Law #2 : Gravity
2cos sin
cos sin 0
sin
M m x ml u
x l g
y x l
&& &&&
&&&&
0
M m x ml u
x l g
y x l
&&&&
&&&&
linearize
Page 55
easy
hard
Law #1 : Mechanics
Law #2 : Gravity
2cos sin
cos sin 0
sin
M m x ml u
x l g
y x l
&& &&&
&&&&
0
M m x ml u
x l g
y x l
&&&&
&&&&
linearize
Page 56
Crashes
can be
made rare
with active
control.
Page 57
hard harder
Easy to prove using simple models.
Why? vision
Act
delay
Law #3 : Light
0
y x
M m x ml u
x l g
l
&&&&
&&&&
Page 58
eye vision
slow
Act
delay
Control
l
1p
l
noise error
E
T jN
Page 59
Universal laws and architectures: brains, bugs, networks, physiology,
grids, medicine, wildfire, turbulence, literature, fashion, dance, earthquakes,
art, music, Lego, buildings, cities vision
Act
delay + Neuroscience
Balancing
an inverted
pendulum
Mechanics+
Gravity +
Light +
2 2
0
1ln
ln
pT j d
p
z pp
z p
Page 60
1
2
4
8
.1 1
Length l (meters)
.05 .5 .2
1p
l
Fragility
p .3s
2 2
0
1 2ln
pT j d p
p
Law #4 :
.1s
Page 61
1
2
4
8
.1 1
Length l (meters)
.05 .5 .2
1p
l
Fragility
p .3s
2 2
0
1 2ln
pT j d p
p
Law #4 :
.1s
Shorter
Page 62
1
2
4
8
.1 1
Length l (meters)
.05 .5 .2
1p
l
Fragility
p .3s
2 2
0
1 2ln
pT j d p
p
Law #4 :
.1s
Page 63
hard harder hardest!
Easy to prove using simple models.
What is sensed matters.
Why?
Page 64
hard harder hardest!
What is sensed matters.
Unstable poles Unstable zeros
0l l0l l
Page 65
hardest!
0l l
2 2
0
l1
n n2
lz pp
T j dp p
pz
0l l
Page 66
1
2
4
8
.1 1 .05 .5 .2 Length (meters)
Fragility
lnz p
p
p
z
.1s
2 2
0
l1
n n2
lz pp
T j dp p
pz
.3s
0 1l l
0l l
Page 67
Sense (Measure
Length
l0, m)
Length l to CoM, m
0.2 0.4 0.6 0.8 1
0.4
0.6
0.8
1
1.2
1
1.5
2
2.5
3
3.5 easy
hard
hard
robust
(easy)
fragile
(hard)
Page 68
Sense (Measure
Length
l0, m)
Length l to CoM, m
0.2 0.4 0.6 0.8 1
0.4
0.6
0.8
1
1.2
1
1.5
2
2.5
3
3.5
hard hard
robust
(easy)
hard!
0l l
easy
Page 69
vision
Act
delay
2 2
0
1ln
ln
pT j d
p
z pp
z p
Holds for all controllers.
Like Turing, a “law” about intrinsic problem difficulty.
Page 70
Select instabilities in biology
Working backwards
• Money/finance/lobbyists/etc
• Society/agriculture/weapons/etc
• Bipedalism
• Maternal care
• Warm blood
• Flight
• Mitochondria
• Translation (ribosomes)
• Glycolysis (see 2011 Science paper)
Page 71
Chandra, Buzi, and Doyle
UG biochem, math, control theory
Most important paper so far.
Page 72
easy
hard
Law #1 : Mechanics Chemistry
Law #2 : Gravity Autocatalysis
Page 73
k
z p
z p
10 -1
10 0
10 1 10
0
10 1
too
fragile
complex
No tradeoff
expensive
fragile
2 2
0
1ln
ln
zS j d
z
z p
z p
Law #1 : Chemistry Law #2 : Autocatalysis
Law #3:
Page 74
1
2
4
8
.1 1 .05 .5 .2
0l l
0 1l l
Length (meters)
Fragility
lnz p
p
p
z
.1s
k
z p
z p
10 -1
10 0
10 1 10
0
10 1
too
fragile
complex
No tradeoff
expensive
fragile
eye vision
Act
slow
VOR
fast
Act
delay
Balancing
an inverted
pendulum Control
Page 75
1
2
4
8
.1 1 .05 .5 .2
0l l
0 1l l
Length (meters)
Fragility
lnz p
p
p
z
.1s
Slow
Flexible
Fast
Inflexible
General Special k
z p
z p
10 -1
10 0
10 1 10
0
10 1
too
fragile
complex
No tradeoff
expensive
fragile
Universal
laws
Page 76
1
2
4
8
.1 1 .05 .5 .2
0l l
0 1l l Fragile
.1s
Slow
Flexible
Fast
Inflexible
General Special
k10 -1
10 0
10 1 10
0
10 1
expensive
fragile
Slow
Page 77
Slow
Flexible
Fast
Inflexible
Robust
Fragile
Slow
Flexible
Fast
Inflexible
General Special
Universal
laws
Page 80
1
2
4
8
.1 1 .05 .5 .2
0l l
0 1l l Fragile
.1s
k10 -1
10 0
10 1 10
0
10 1
expensive
fragile
Page 81
1
2
4
8
.1 1 .05 .5 .2
0l l
0 1l l Fragile
.1s
Slow
Flexible
Fast
Inflexible
General Special
k10 -1
10 0
10 1 10
0
10 1
expensive
fragile
Slow
Page 82
Universal laws
and architectures: nets, grids, brains, bugs, bodies, doctors, fire, fashion,
art, turbulence, music, buildings, cities, earthquakes, planes, running, throwing,
Synesthesia, spacecraft, statistical mechanics
Slow
Flexible
Fast
Inflexible
General Special
Page 83
Slow
Flexible
Fast
Inflexible
General Special
Universal Turing Machine
Page 84
Fast
Slow
Flexible Inflexible
Apps
OS
HW
Apps
OS
Hardware
Digital
Lumped
Distributed
Tech implications/extensions
Page 85
Act
Sense
Apps
OS
Hardware
Digital
Lumped
Decide
Slow
vision
Act
delay
Page 86
vision
Act
delay
+ Neuroscience
Completing the story
Balancing
an inverted
pendulum
Mechanics+
Gravity +
Light +
2 2
0
1ln
ln
pT j d
p
z pp
z p
Page 87
Speed and
flexibility
tradeoffs
Concrete case studies? Theorems ?
Fast
Slow
Flexible Inflexible
fragile
robust
Page 88
Slow
Flexible
Fast
Inflexible
Robust
Fragile
Cheap
Expensive
Efficient vs Robust
Page 89
Slow
Flexible
vision
eye vision
Act slow
delay
Fast
Inflexible
VOR
fast
Page 90
eye vision
Act slow
VOR
fast
Act
delay
Balancing
an inverted
pendulum Control
Page 91
1
2
4
8
.1 1 .05 .5 .2
0l l
0 1l l
Length (meters)
Fragility
lnz p
p
p
z
.1s
Slow
Flexible
Fast
Inflexible
General Special
eye vision
Act
slow
VOR
fast
Act
delay
Balancing
an inverted
pendulum Control
Page 92
Slow
Flexible
vision
eye vision
Act slow
delay
Fast
Inflexible
VOR
fast
3D + motion
color
vision
Slowest
Page 93
Slow
Flexible
vision
eye vision
Act slow
delay
Fast
Inflexible
VOR
fast
B&W (luminence only): 3D, motion, and action
3D + motion
Page 94
This is pretty good. Stare at the intersection
Page 96
This is pretty good. Stare at the intersection
Page 98
Slow
Flexible
vision
eye vision
Act slow
delay
Fast
Inflexible
VOR
fast
3D + motion
color
vision
Slowest
Page 99
Slow
Flexible
vision
Fast
Inflexible
VOR
color
vision
Seeing is dreaming
Page 100
- 3d+motion
B&W
slow
VOR
fast
Fast
Slow
Flexible Inflexible
slower
Objects
Mixed
Page 102
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
Page 103
- 3d+motion
B&W
slow
VOR
fast
Fast
Slow
Flexible Inflexible
Faces
slower
Objects
Mixed
Not sure how to draw this…
Page 104
vision
eye vision
Act slow
delay
VOR
Slow
Flexible
Fast
Inflexible
Illusion
Highly
evolved
(hidden)
architecture
Seeing is dreaming
Page 105
3D +time
Simulation
+ complex
models
(“priors”)
Seeing is dreaming
Conscious perception
Conscio
us
pe
rcep
tio
n
errors
Prediction
&Control
Page 106
Fast
Costly
Slow
Cheap
Flexible Inflexible
General Special
Other dimensions?
Page 107
1x
2x
3x
Model?
• 1 dimension, 4 states?
• Other 2 dimensions?
• New issues arise
Page 108
eye vision
Act
delay
Control
noise error
E
T jN
1x
2x
3x
Page 109
eye vision
Act
delay
Control
noise error
E
T jN
1x
2x
3xeasy
easy
hard
Page 110
eye vision
Act
delay
Control
noise error
E
T jN
1x
2x
3xeasy
easy
hard
Page 111
Main lessons
• Theory: hard limits on closed loop
performance, aggravated by
– Instability (unstable poles)
– Delay
– Unstable zeros
• Neuroscience specific
Page 112
Instabilities in technology
• Efficiency
• Autocatalysis
• The future
Page 113
Select instabilities in biology
Working backwards
• Money/finance/lobbyists/etc
• Society/agriculture/weapons/etc
• Bipedalism
• Maternal care
• Warm blood
• Flight
• Mitochondria
• Translation (ribosomes)
• Glycolysis (see 2011 Science paper)
Page 114
sensor
controls
external
disturbances
heart rate
ventilation vasodilation
coagulation
inflammation
digestion
storage
…
errors
O2
BP pH
Glucose
Energy store
Blood volume
…
infection
trauma energy
Homeostasis
internal noise
heart beat
breath
Page 115
control feedback
RNA
mRNA
RNAp
Transcription
Other
Control
Gene
Amino
Acids
Proteins
Ribosomes
Other
Control
Translation
Metabolism Products
Signal transduction
ATP
DNA
New
gene
Sensory Motor
Prefrontal
Striatum
Reflex
Software
Hardware
Digital
Analog
Horizontal
Gene
Transfer
Horizontal
App
Transfer
Horizontal
Meme
Transfer
Page 117
Linux kernel map OS
Page 121
User
interface System Processing Memory Storage Network
User space
interfaces
Virtual
Bridges
Logical
Device
control
Hardware
Interfaces
Hardware
Linux kernel map
Page 122
SW Layer
SW Layer
SW Layer
SW Layer
SW Layer
SW Layer
User
interface System Processing Memory Storage Network
User space
interfaces
Virtual
Bridges
Logical
Device
control
Hardware
Interfaces
Hardware
Page 123
User space
interfaces
Virtual
Bridges
Logical
Device
control
Hardware
Interfaces
Hardware
User
interface System Processing Memory Storage Network
Functions
use all layers
User peripherals
I/O CPU Memory Disk controller
Network controller
Page 124
Outfit
Cloth
Thread
Fiber
Cloth
Thread
Garment
Cloth
Thread
Fiber
Cloth
Thread
Garment
Cloth
Thread
Fiber
Cloth
Thread
Garment
Page 125
Outfit
Cloth
Thread
Fiber
Cloth
Thread
Garment
Cloth
Thread
Fiber
Cloth
Thread
Garment
Cloth
Thread
Fiber
Cloth
Thread
Garment
Page 128
robust
costly
fragile
cheap
Page 129
Outfit B
ase
Insu
latio
n
Sh
ell
Body Environment
Layering & Modularity 1.0
Page 130
Outfit
Cloth
Thread
Fiber
Cloth
Thread
Garment
Cloth
Thread
Fiber
Cloth
Thread
Garment
Cloth
Thread
Fiber
Cloth
Thread
Garment
Page 131
Cloth
Thread
Fiber
Cloth
Thread
Garment
Cloth
Thread
Fiber
Cloth
Thread
Garment
Cloth
Thread
Fiber
Cloth
Thread
Garment
Architecture, Layering & Modularity 2.0
Page 132
OS
Layered architectures
Minimal
diversity
and
change
Apps
OS
HW
Page 133
Software
Hardware
Digital
Analog
Cloth
Thread
Fiber
Cloth
Thread
Garments
Components
Organized
Components
Organized
What is the “true functional unit?”
Page 134
Virtual machines
Software
Hardware
Digital
Analog
Cloth
Thread
Fiber
Cloth
Thread
Garments
Components
Organized
Components
Organized
Organized Software Garments
Implementations
Page 135
Virtual machines
Digital
Analog
Thread
Fiber
Cloth
Thread Components
Organized
Implementations
Organized Software Garments
Digital Cloth Organized
Page 136
Virtual Outfit
Garment Garment Garment
Page 137
Physical connectivity
Fiber Fiber Fiber
Page 138
Outfit
Cloth
Thread
Fiber
Cloth
Thread
Garment
Cloth
Thread
Fiber
Cloth
Thread
Garment
Virtual
Physical
Page 139
Ga
rme
nt
Ga
rme
nt
Ga
rme
nt
Outfit
Cloth
Thread
Fiber
Garment
Garm
ent
Cloth
Thread
Fiber
Garment G
arm
en
t
Cloth
Thread
Fiber
Garment
Layering within garments (textiles)
Page 140
Doyle, Csete, Proc Nat Acad Sci USA, JULY 25 2011
For more info Most accessible
No math References
Page 141
Slow Apps
HW
Implementation
Control of function
Flexible
General
Fast
Inflexible
Special
Page 142
Fast
Slow
Flexible Inflexible
General Special
Apps
OS
HW
Horizontal
App
Transfer
Horizontal
HW
Transfer
Page 143
Fast
Costly
Slow
Cheap
Flexible Inflexible
General Special
Layered
Architecture
Apps
OS
HW
Digital
Lumped
Distrib.
Any layer needs
all lower layers.
Page 144
Digital
Lumped
Distrib.
You can control function
in hardware alone.
Lumped
Distrib.
Distrib.
Hardware
Digital
Lumped
Distributed
Page 145
- 3d+motion
B&W
slow
VOR
fast
Fast
Slow
Flexible Inflexible
Faces
slower
Objects
Mixed
Not sure how to draw this…
Page 146
Sensory Motor
Prefrontal
Striatum
Slow
Flexible
Learning
Ashby & Crossley
Slow
Reflex (Fastest,
Least
Flexible)
Extreme heterogeneity
Page 147
Sensory Motor
Prefrontal
Striatum
Fast
Fast
Inflexible
Slow
Flexible
Reflex (Fastest,
Least
Flexible)
Ashby & Crossley
Learning can
be very slow.
Page 148
Sense
Universal tradeoffs?
Fast
Slow
Flexible Inflexible
Motor
Prefrontal
Fast
Reflex
Learning can
be very slow.
Evolution is
even slower.
Page 149
Sense
Fast
Slow
Flexible Inflexible
Motor
Prefrontal
Fast
Reflex
Apps
OS
HW
Page 150
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??
Constrained Brain
Page 151
0. HGT
1. DNA repair
2. Mutation
3. DNA replication
4. Transcription
5. Translation
6. Metabolism
7. Signal transduction
8. …
Apps
OS Hardware
Digital
Lumped
Distributed
Any layer
needs all
lower layers.
control feedback
RNA
mRNA
RNAp
Transcription
Other
Control
Gene
Amino
Acids
Proteins
Ribosomes
Other
Control
Translation
Metabolism Products
Signal transduction
ATP
DNA
New
gene
Page 152
control feedback
RNA
mRNA
RNAp
Transcription
Other
Control
Gene
Amino
Acids
Proteins
Ribosomes
Other
Control
Translation
Metabolism Products
Signal transduction
ATP
DNA
New
gene
Fast
Costly
Slow
Cheap
Flexible Inflexible
General Special
HGT DNA repair Mutation DNA replication
Transcription Translation
Metabolism Signal…
Layered
Architecture
Page 153
Slow
Fast
Flexible Inflexible
General Special
Apps OS HW Dig. Lump. Distrib.
OS HW Dig. Lump. Distrib.
Digital Lump. Distrib.
Lumped Distrib.
Distrib.
HGT DNA repair Mutation DNA replication Transcription Translation Metabolism
Signal
Sense Motor
Prefrontal
Fast
Reflex
vision
VOR
Page 154
Slow
Fast
Flexible Inflexible
General Special
Apps OS HW Dig. Lump. Distrib.
OS HW Dig. Lump. Distrib.
Digital Lump. Distrib.
Lumped Distrib.
Distrib.
HGT DNA repair Mutation DNA replication Transcription Translation Metabolism
Signal
Sense Motor
Prefrontal
Fast
Reflex
VOR
Many systems/organisms only use lower layers
Page 155
Lumped Distrib.
Distrib.
Metabolism Signal
Sense Motor Fast
Reflex
VOR
Many systems/organisms only use lower layers
Analog
Red blood cells
Most metazoans
Page 156
Slow
Fast
Flexible Inflexible
General Special
Apps OS HW Dig. Lump. Distrib.
OS HW Dig. Lump. Distrib.
Digital Lump. Distrib.
Lumped Distrib.
Distrib.
Ideal?
Sense Motor
Prefrontal
Fast
Reflex
vision
VOR
Can a new architecture beat the tradeoff?
Page 157
Slow
Fast
Flexible Inflexible
General Special
Apps OS HW Dig. Lump. Distrib.
OS HW Dig. Lump. Distrib.
Digital Lump. Distrib.
Lumped Distrib.
Distrib.
Sense Motor
Prefrontal
Fast
Reflex
vision
VOR
Be careful what you wish for…
Page 158
Apps
OS
HW
Digital
Lumped
Distrib.
OS
HW
Digital
Lumped
Distrib.
Digital
Lumped
Distrib.
Lumped
Distrib.
Distrib.
VOR
Slow
Fast
Flexible Inflexible
General Special
Be careful
vision
?
Page 159
control feedback
RNA
mRNA
RNAp
Transcription
Other
Control
Gene
Amino
Acids
Proteins
Ribosomes
Other
Control
Translation
Metabolism Products
Signal transduction
ATP
DNA
New
gene
Sensory Motor
Prefrontal
Striatum
Reflex
Software
Hardware
Digital
Analog
Horizontal
Gene
Transfer
Horizontal
App
Transfer
Horizontal
Meme
Transfer
Page 160
control feedback
RNA
mRNA
RNAp
Transcription
Other Control
Gene
Amino
Acids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transduction
ATP
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
Clinical issue: antibiotic resistance?
Page 161
control feedback
RNA
mRNA
RNAp
Transcription
Other Control
Gene
Amino
Acids
Proteins
Ribosomes
Other Control
Translation
Metabolism Products
Signal transduction
ATP
DNA
New
gene
Sensory Motor
Prefrontal
Striatum
Reflex
Software
Hardware
Digital
Analog
Horizontal Gene
Transfer
Horizontal App
Transfer
Horizontal Meme
Transfer
Page 162
control feedback
RNA
mRNA
RNAp
Transcription
Other
Control
Gene
Amino
Acids
Proteins
Ribosomes
Other
Control
Translation
Metabolism Products
Signal transduction
ATP
DNA
Sensory Motor
Prefrontal
Striatum
Reflex
Software
Hardware
Digital
Analog
Enable
“vertical”
innovation
Mechanisms for
Horizontal Transfer
Page 163
control feedback
RNA
mRNA
RNAp
Transcription
Other
Control
Gene
Amino
Acids
Proteins
Ribosomes
Other
Control
Translation
Metabolism Products
Signal transduction
ATP
DNA
New
gene
Horizontal
Gene
Transfer
What can
go wrong?
Page 164
control feedback
RNA
mRNA
RNAp
Transcription
Other
Control
Gene
Amino
Acids
Proteins
Ribosomes
Other
Control
Translation
Metabolism Products
Signal transduction
ATP
DNA
New
gene
Sensory Motor
Prefrontal
Striatum
Reflex
Software
Hardware
Digital
Analog
Horizontal
Gene
Transfer
Horizontal
App
Transfer
Horizontal
Meme
Transfer
What can
go wrong?
Page 165
Horizontal
Bad Gene
Transfer
Horizontal
Bad App
Transfer
Horizontal
Bad Meme
Transfer
Parasites
&
Hijacking
Fragility?
Exploiting
layered
architecture
Virus
Virus
Meme
Page 166
Fast
Costly
Slow
Cheap
Flexible Inflexible
General Special
Layered
immunity
Adaptive
OS?
Innate
Vaccination ≈ horizontal transfer?