Session 15: Episode 4 —
21st Century global brains and humano-technical cyborgs
William P. Hall President Kororoit Institute Proponents and Supporters Assoc., Inc. - http://kororoit.org [email protected] http://www.orgs-evolution-knowledge.net
Access my research papers from Google Citations
An important new paper on the relationship between the adjacent possible and the relevant next
Referring back to the role of the “adjacent possible” in “crystallizing spacetime” see Favareau, D.F. 2015. Creation of the relevant next: How living systems capture the power of the adjacent possible through sign use. Progress in Biophysics and Molecular Biology (2015), http://dx.doi.org/10.1016/j.pbiomolbio.2015.08.010.
Stuart Kauffman's notion of the Adjacent Possible as an organizing principle in nature shares much with Charles S. Peirce's understanding of the universe as an ever unfolding ‘process ontology’ of possibility space that is brought about through the recursive interaction of genuine possibility, transiently actualized order, and emergent (but never fully deterministic) lawfulness. Proceeding from these three fundamental categories of becoming-as-being, Peirce developed a complimentary logic of sign relations that, along with Jakob von Uexküll's action-as-meaning-imprinting Umwelt theory, informs work currently being undertaken under the aegis of Biosemiotics.
This paper highlights the deep affinities between Kauffman's notion of the Adjacent Possible and Biosemiotics' hybrid Peircean/Uexküllian “sign” concept, by which living systems - both as individuals and in the aggregate (i.e., as co-actors, communities and lineages) - “capture” relevant aspects of their relations with the immediately given Adjacent Possible and preserve those recipes for future interaction possibilities as biologically instantiated signs.
By so doing, living systems move into the Adjacent Possible by “collapsing the wave function” of possibility not just probabilistically, but guided by system-internal values arising from previously captured sign relations that are biologically instantiated as replicable system biases and generative constraints. The influence of such valenced and end-directed action in the world introduces into the universe the phenomenon of the Relevant (and not just deterministic, or even stochastic) Next.
The paper argues that organisms live out their lives perpetually confronted with negotiating the omnipresent Relevant Next, and are informed by the biological capture of their (and their lineage's) previous engagements in doing so. And because that “capture” of previous agent-object-action relationships are instantiated as biological signs for the guidance of the organism, not only are “successful survival strategies” within a given possibility space captured (as in traditional accounts of Natural Selection), but captured as well within those signs are the entire complement of previously untaken but still veridical real-world possibility spaces that are inseparably ‘entangled’ with that sign, and just awaiting exploration by the organism.
Thus, while all action in the universe is both current-context dependant and next-context creating, the emergence of ever-more complex semiotic capabilities in organisms has expanded the possibility space of immediate-next-action in the world exponentially, and has brought into being not a pre-given, singly end-directed ordered world, but an emergent, many ends-directed world of promiscuous, unforeseeable and interacting telos. The goal of Biosemiotics is to understand and to explore this world 2
Tonight
In the last two sessions, in an “Interlude”, I summarized theories of life, cognition, and knowledge based on Maturana and Varela’s autopoietic definition of life and Karl Popper’s evolutionary epistemology (the Favareau paper extends this)
In this session I explore the disruptive convergence of technology and human biology and technology.
– Convergence forms hybrid individuals with both biological and technological capabilities
– Gaining pace since Application Holy Wars project started 15 years ago
3
Extending Human Cognition and Emergence of Humano-Technical Cyborgs Moore’s Law still at work – clouds, pipes, devices, and apps. Human Computer Interfaces (HCI) – sensing, processing, augmenting cognition The next steps in merging human and computer cognition
Intimately wearable interfaces Implanted/embodied human-machine interfaces
Moore’s Law and its implications for embodied interfaces What does it mean to be human?
Autopoietic boundaries and cognition Neural basis for self-consciousness Human evolution in several dimensions
Introducing Episode 4
Convergence of cellular functional organization – Similar to endosymbiotic origin of eukaryotic cell
– Host: cytoplasmic motility, chromosomal inheritance, sexual isolation
– Symbionts: Bacteria – organic activation driven by coupled oxidation reactions
Cyanobacteria – chemosynthesis driven by photonically activated electrons
Technological convergence – Refers to the situation in technological evolution over time where a single
technology (i.e., tool) subsumes functions that were previously provided by several more specifically specialized tools.
– One of the most extreme examples is the “smartphone”
Cyborg: Posthuman convergence of human biology and cognition with various sensor, effector, cognitive, and communication technologies
– Radically changes the nature of humanity
[first order autopoietic systems – emergence of “structural” knowledge]
(organelles - endosymbionts – Hall 1966 membranes, filaments, microtubules)
eukaryotic cells - meiotic recombination and random assortment
prokaryotic cells - “genetic knowledge”
[second order autopoietic systems (multicellular)]
Moore’s Law at work
—
Human’s use of technology may have
changed as much in the last 15 years as it did in the previous 50 years
— We are now interacting with machines, knowledge and with other people via personal connections into computer mediated virtual worlds in ways
that were unimaginable a couple of decades ago
Moore’s Law applies to much more than transistors
6
Calculations per second per $1000
Hyperdeflating storage costs for digital data and the rise of the “Cloud”
What was economically impossible in 2000 is now essentially free in 2015
My personal storage: – 38 GB on Google Drive for $5.00
per year
– 18 GB on Dropbox for no cost
– Essentially my entire working library and working history now lives redundantly in the cloud
7
Aug. 2011 Aug. 2013 Aug. 2015
First 1 TB / month 0.1400 0.0950 0.0300
Next 49 TB / month 0.1250 0.0800 0.0295
Next 450 TB / month 0.1100 0.0700 0.0290
Next 500 TB / month 0.0950 0.0650 0.0285
Next 4000 TB / month 0.0800 0.0600 0.0280
Over 5000 TB / month 0.0550 0.0550 0.0275
Amazon S3 Price per GB Storage in $
http://www.martinhilbert.net/WorldInfoCapacity.html/
Hyperdeflating costs for bandwidth and the movement of data
8
Bandwidth cost-performance (1999-2012) Source dupress.com
Transport becomes a function of some other service
Wired vs wireless - Tracking wireless pricing difficult!
Who maintains a service that costs nothing to provide?
- Pay to access the people who use it
- Video on demand
Moore’s Law and converging devices
9
Mobile phones: – 1989: Motorola’s state of the art MicroTAC 9800x
Cost US$2,495 Weight 349 gms Memory for a few phone numbers
– 2015: Samsung Galaxy S6 edge+ 64 GB Cost US $780 (25/08/2015) Weight 153 gms 64 gms Touch screen 2560px by 1440px Network phone/data: 2G, 3G, 4G Connectivity: Wi-Fi; Bluetooth Talk time – up to 20 hours; Standby – up to 557 hours Audio recording/play: Play up to 57 hrs Cameras: Rear – 16 megapixel; Front – 5 megapixel; LED flash Video: UHD 4K (3840px 2160px) @ 30fps; Play up to 13 hrs Sensors: Accelerometer, Barometer, Fingerprint Sensor, Gyro Sensor,
Geomagnetic Sensor, Hall (magnetic field) Sensor, HR Sensor, Light Sensor, Proximity Sensor
Smartphone apps determine what the convergent systems can do (sample only)
Voice communication (i.e., the original telephone function) – wireless: mobile telephony (1946), cell phone (1956), first hand-held mobile (1973).
– internet telephony: development of Voice over IP (1973) – first app (1994).
– Google Voice (2005/2007 - free): Convergent features include voicemail, call forwarding, texting, call history, conference calling, screening, voice transcription to text, etc.
Text communication – email: TCP/IP (1982); Gmail (Apr 2005 / Feb 2007 - free). Convergent features include search, cloud storage, instant
messaging (voice & text), calendar functions
Portable media players/media libraries/sales of media, software, etc… – Windows Media Player (1991 - free); iTunes (1999/2001). Convergent features include playing, downloading, saving, and
organizing digital music and video files, connecting to iTunes Store to purchase/download media & app software for iPads and iPhones
– Amazon Kindle (2007): Tablet sized book reader using E Ink electronic paper, evolving convergence
Photography & video-related applications – Picassa (2002 - free); Panoramio (2005 - free); Geolocated photos converging with Google Earth and Google Maps (e.g.,
Natureshare)
– YouTube (2005 - free): Video sharing and upload (also manages sound only files)
Cloud storage and file sharing – Napster (1999); BitTorrent protocols (2001); DropBox (2008 – free)
Business and office tools via cloud for smartphones – Google Docs/Google Drive (2007 - free): Cloud-based authoring apps working via browser, plus cloud storage, searching, and
file sharing.
Geospatial – Google Earth (2001/2005 - free); Google Maps (2005 – free)
Social – Chat rooms (~1980); Email groups/listservers (Majordomo 1992); Meetup (2001): LinkedIn (2003); Facebook (2004);
Twitter (2006)
Knowledge construction, sharing and broadcasting – Wikis (1994); Wikipedia (2002); Blogs (late 1990s); Google Sites combines blogging and wiki functions with file sharing 10
Human-Computer Interfaces (HCI)
—
Convergence of biology and technology
Responding to the world via the OODA Loop: Perception
12
Loomis 2003
Francis et al. (2009)
Processing Paradigm(may include W3)
Decision
Medium/Environment Autopoietic system
World State 1
PerturbationTransduction
Observation MemoryClassification
Evaluation
Synthesis
AssembleResponse
Internal changes
Effect action
Effect
Time
World State 2
IterateObserved internal changes
World 1(physics & dynamics)
World 2Internal environment(physiology, neural connections + memory)
Codified knowledge
World 3
Orientation
Observe perturbationscaused by prior actionson the environment
Augmenting cognition via technological interfaces
— Next steps in merging human and computer
cognition
Intimately wearable interfaces
Smartphones plugged into people – Hand-held – In pockets, in ears – Enhancements
Wireless earpieces Voice recognition/synthesis
– Limited by auditory bandwidth
Google Glass – Reality “augmented” with overlays – “head-up display” (input to the human eye),
camera suitable for lifelogging, speakers (audio input), microphones, processor, memory, touch pad on the earpiece, modes, sensors (gyroscopes, accelerometers and a compass), and instant data access
– Data should be accessible “so fast that you feel you know it” – Privacy & safety issues may have impeded takeup
Smart lenses? (Parviz 2009); Google smart contact lens Smartwatch convergent apps as per smartphone as enabled by
Moore’ 14
Implanted embodied human-machine interfaces
Bionic ear
15
Bionic eye
Output transducers (attach to peripheral nerves) - Control for bionic limbs, etc.
- In principal, can control all kinds of external effectors via internet of things
Moore’s Law and its implications for embodied interfaces
— Hardware, software & wetware
Nanowiring able to interface single nerve cells (hard but doable) – Material: metal, semiconductor, carbon nanotubes, graphene
– Adjust chemistry to target specific types of nerve cells
Understand wetware – Meso-scale interconnections
– Micro-scale mapping Hippocampal region (central indexing sysem)
Cortical columns
Converging technology and biology
17
Functional organization of long-term memory
18
Long-term memory
Declarative (explicit) Procedural (implicit)
Episodic Semantic
Spatial
Know how Body memory
What happened when? Temporal indexing
Abstracted facts, meanings, concepts, models of external world i.e., our “scientific” knowledge of how the world works Classification indexing
Cognitive map of the experienced environment What are the landmarks? Where I am on the map? What happened here? What happened there? Where is home? What else do I know about here and there? Spatial monitoring & feedback Navigating mapping spatial indexing
Short-term memory
Ongoing interactions in the world
Performance monitoring & feedback
Hippocampal region
Mapping, understanding and simulating wetware processing
Challenges – Neural complexity
different kinds of neurons,
variety of synapses
– Scale : the human cortex has tens of billions of neurons and a quadrillion or so synapses.
– Interconnectivity A neuron may share synapses with 10,000 up to 100,000 other neurons.
Axons may fan out to an average of 10,000 different destinations.
– Plasticity : Memory and learning imply substantial plasticity
– Power consumption : Human brain dissipates around 25 watts!
Computational approaches – Simulation on modular supercomputers
– Emulation using custom multi-cpu chips Current (2012) address structure supports 65,000 chips
Able to emulate 1 BN neurons
19
Neuromorphic simulation
20
Kim et al. 2015
Standard von Neumann architecture many orders of magnitude too energy intensive.
Neuromorphic architecture – Memrister crossbar
arrays
– Spiking “neurons”
– Energy consumption approaches biological values
Example: IBM TrueNorth chip (now) – 5.4 x 109 transistors
– 256 x 106 synapses
– 106 “neurons”
– Power density 20 mW/cm2 ~ 10-4 that of a normal von Neumann chip
– Steps 16 chips, 16 million neurons; 4 billion synapses
Mouse brain (48 x 106 neurons – chip module)
4,096 Chips; 4 billion neurons; 1 trillion synapses
Human 10 billion neurons; 100 trillion synapses
Cognitive convergence between wetware and hardware/software devices
Brain computer interfaces already being implemented to control prosthetics for physical/neurological impairments (pacemakers, bionic ears, bionic eyes, artificial limbs)
– To be effective these the interfaces already involve substantial (cognitive) computational processing
– Effectors – such as the external manipulators do not even require physical connection (other than electronic) to the body E.g., the connection can be via wi-fi for both control and feedback
Moore’s law suggests that the bandwidth and capabilities of the sensors and effectors will only increase with time
There is no reason why such bionic sensors and effectors cannot interface equally well with simulated brains
– Emulated mouse brain in an emulated mouse body (very soon)
– Emulated cat brain in an emulated cat body
– Emulated primate brain in an emulated primate body
– Emulated human brain in an emulated human body
– Etc….. 21
What does it mean to be post-human?
The autopoietic post-human human entity consists of the self-conscious entity and all of its sensory, cognitive, and effector components
– The post-human entity today is no longer bounded by its skin
– It is vastly more powerful and plastic than it was, say, 100 years ago
– The bandwidth of its interconnectivity with other post-human entities is vastly larger than it was even 20 years ago
“Consciousness” of decision & action follow action by milliseconds but may “predecide” future decisions to anticipate real world requirements
22 "comparator model" of agency (after Morsella et al. 2010
Human volition cycle (after Haggard et al. 2008)
Next session begins Episode 5 - Extending social cognition and emergence of socio-technical organizations
23
Begins the last, largest and most complex episode in my fugue – Explores & reconstructs the emergence and evolution of humanity
and human organizations from a family of tool-using apes over the last 4-5 million years
– Biological rather than a technological point of view
Part 1 of the episode sets out – Basic concepts of evolutionary and comparative biology used in the
reconstruction – Summarizes the material (i.e., fossil, genomic & archaeological)
evidence
Extending Social Cognition and Emergence of Socio-Technical Organizations
Introduction
Material evidence: what we think we know about hominin evolution
What fossils tell us about our pedigree and relationships
• Homo emerges and crosses the continents
What comparative genomics tells us about our genealogy
• Denisovans
• Neanderthals
What ancient tool-kits tell us about our ancestors
