COGNITIVE MESH
as Artificial Intelligence 2.0:
Associative & Abstracting
Blockchain-based
Biologically-inspired
Competitive and Self-evolving
Dr Maxim Orlovsky, MD, PhD
orlovsky@bicalabs.org
Modern approach to #AI has grown from
research of 1950s-60s and is outdated
➔ Supervised Learning
No ad hoc concept creation and
feature extraction without previous
knowledge
➔ Not a Brain
A brain region can’t mimic the whole
➔ Lack of Multiplicity
Leads to noncompetitive
development
➔ Enclosed
There is no reflection and cross-
agent knowledge exchange
MODERN
AI
IS
OUTDATED
SUPERVISED
LEARNING
“I see only what I know” problem
CONVOLUTIONAL
NEURONETS
REPRESENT
JUST A CORTEX ZONE,
NOT THE BRAIN
EVOLUTIONARY EVOLVEMENT
HAPPENS WHEN
SELECTION ACTS ON A
DIVERGING POPULATION
WHILE
MODERN NEURONETS
ARE SINGULAR
INFORMATION TRANSFER:
● species exchanges genes via viruses
● cells emit chemical and electric signals
● animals communicate via mimics and emotions
● humans — via language
WHAT ABOUT #AI?
SELF-REFLECTION
IS THE OTHER SIDE OF
COMMUNICATIONS
CAN #AI
REFLECT
ITSELF?
THESE PROBLEMS ARE
ADDRESSED BY
BLOCKCHAIN-BASED
MULTI-AGENT AI
SHARING “ASSOCIATIVE MEMORY GRAPH”
MULTI-AGENT
Multiple artificially-intelligent
systems
(like neural networks)
sharing the same internal
structure
DIVERGING
Agents adapt to the
information given by adjusting
internal parameters
(like weights etc) with some
“noise salt”
BLOCKCHAIN OF
ABSTRACTIONS AND
ASSOCIATIONS (MEMORY)
Agents share results of their
cognition as symbols
(abstractions) and their
associations into a blockchain
of a “shared knowledge” and
memory.
Entries are signed by a
“proof of cognitive work”
REFLECTION
Abstract symbols in blockchain
represent a sort of “written
language”
Agents can reflect upon their
own cognitive work by
interacting with their previous
records in blockchain
COMPETITION
AND SELECTION
Agents “economically
compete” by their cognitive
efficiency employing internal
blockchain cryptocurrency
“paying” for the usage of
cognitive information from
each other in form of “smart
contracts”
BLOCKCHAIN WILL GIVE TO
MULTI-AGENT AI
the same thing that written language gave to humans:
MEANS OF
COGNITIVE EVOLUTION
https://medium.com/bica-labs/how-blockchain-relates-to-artificial-intelligence-f0111f39afc9
INFORMATION TRANSFER
REQUIRES
CONSENSUS AND
PERSISTENCE
CONSENSUS:
each agent gets the same
sense of a signal as others
– DNA translation will be the same for bacteria as for a human being
– written text is understood in the same way by all native speakers
https://medium.com/bica-labs/how-blockchain-relates-to-artificial-intelligence-f0111f39afc9
PERSISTENCE:
the sense of the signal does
not diminish with time
– DNA in chromosomes do not change during the live of the most cells
– Egyptian hieroglyphs could still be understood by researchers
BLOCKCHAIN
IS PERFECT BOTH FOR
CONSENSUS AND
PERSISTENCE
RENDERING MULTI-AGENT #AI AS AN
EVOLVING SELF-ORGANIZED SYSTEM
COGNITIVE MESH
BY BICA LABS AND EVENT HORIZONS
CURRENT STAGE:
– DEVELOPING SPECS AND TECH PROTOTYPES
– SOON: INITIAL PUBLIC OFFERING ON ETHEREUM
HTTP://COGNITIVEMESH.COM
TWITTER: @BICALABS
BICA Labs
Biologically Inspired Cognitive Architectures
orlovsky@bicalabs.org
http://bicalabs.org
@BICALabs
fb.com/BICALabs
linkedin.com/company/BICA-Labs
medium.com/BICA-Labs
Private foundation & task force supporting
disruptive technologies and radical innovations
having potential to change world politics,
economics and society
hello@eventhorizons.org
http://eventhorizons.org
@EvHorizons
fb.com/EvHorizons
linkedin.com/company/Event-Horizons
medium.com/Event-Horizons

Artificial Intelligence 2.0: Cognitive Mesh with Blockchain

  • 1.
    COGNITIVE MESH as ArtificialIntelligence 2.0: Associative & Abstracting Blockchain-based Biologically-inspired Competitive and Self-evolving Dr Maxim Orlovsky, MD, PhD orlovsky@bicalabs.org
  • 2.
    Modern approach to#AI has grown from research of 1950s-60s and is outdated ➔ Supervised Learning No ad hoc concept creation and feature extraction without previous knowledge ➔ Not a Brain A brain region can’t mimic the whole ➔ Lack of Multiplicity Leads to noncompetitive development ➔ Enclosed There is no reflection and cross- agent knowledge exchange MODERN AI IS OUTDATED
  • 3.
    SUPERVISED LEARNING “I see onlywhat I know” problem
  • 4.
  • 5.
    EVOLUTIONARY EVOLVEMENT HAPPENS WHEN SELECTIONACTS ON A DIVERGING POPULATION WHILE MODERN NEURONETS ARE SINGULAR
  • 6.
    INFORMATION TRANSFER: ● speciesexchanges genes via viruses ● cells emit chemical and electric signals ● animals communicate via mimics and emotions ● humans — via language WHAT ABOUT #AI?
  • 7.
    SELF-REFLECTION IS THE OTHERSIDE OF COMMUNICATIONS CAN #AI REFLECT ITSELF?
  • 8.
    THESE PROBLEMS ARE ADDRESSEDBY BLOCKCHAIN-BASED MULTI-AGENT AI SHARING “ASSOCIATIVE MEMORY GRAPH”
  • 9.
    MULTI-AGENT Multiple artificially-intelligent systems (like neuralnetworks) sharing the same internal structure
  • 10.
    DIVERGING Agents adapt tothe information given by adjusting internal parameters (like weights etc) with some “noise salt”
  • 11.
    BLOCKCHAIN OF ABSTRACTIONS AND ASSOCIATIONS(MEMORY) Agents share results of their cognition as symbols (abstractions) and their associations into a blockchain of a “shared knowledge” and memory. Entries are signed by a “proof of cognitive work”
  • 12.
    REFLECTION Abstract symbols inblockchain represent a sort of “written language” Agents can reflect upon their own cognitive work by interacting with their previous records in blockchain
  • 13.
    COMPETITION AND SELECTION Agents “economically compete”by their cognitive efficiency employing internal blockchain cryptocurrency “paying” for the usage of cognitive information from each other in form of “smart contracts”
  • 14.
    BLOCKCHAIN WILL GIVETO MULTI-AGENT AI the same thing that written language gave to humans: MEANS OF COGNITIVE EVOLUTION https://medium.com/bica-labs/how-blockchain-relates-to-artificial-intelligence-f0111f39afc9
  • 15.
  • 16.
    CONSENSUS: each agent getsthe same sense of a signal as others – DNA translation will be the same for bacteria as for a human being – written text is understood in the same way by all native speakers https://medium.com/bica-labs/how-blockchain-relates-to-artificial-intelligence-f0111f39afc9
  • 17.
    PERSISTENCE: the sense ofthe signal does not diminish with time – DNA in chromosomes do not change during the live of the most cells – Egyptian hieroglyphs could still be understood by researchers
  • 18.
    BLOCKCHAIN IS PERFECT BOTHFOR CONSENSUS AND PERSISTENCE RENDERING MULTI-AGENT #AI AS AN EVOLVING SELF-ORGANIZED SYSTEM
  • 19.
    COGNITIVE MESH BY BICALABS AND EVENT HORIZONS CURRENT STAGE: – DEVELOPING SPECS AND TECH PROTOTYPES – SOON: INITIAL PUBLIC OFFERING ON ETHEREUM HTTP://COGNITIVEMESH.COM TWITTER: @BICALABS
  • 20.
    BICA Labs Biologically InspiredCognitive Architectures orlovsky@bicalabs.org http://bicalabs.org @BICALabs fb.com/BICALabs linkedin.com/company/BICA-Labs medium.com/BICA-Labs Private foundation & task force supporting disruptive technologies and radical innovations having potential to change world politics, economics and society hello@eventhorizons.org http://eventhorizons.org @EvHorizons fb.com/EvHorizons linkedin.com/company/Event-Horizons medium.com/Event-Horizons