SlideShare une entreprise Scribd logo
1  sur  44
Télécharger pour lire hors ligne
How AI benefits from Blockchain
and Game Theory with 

Scalable Censorship-resistant
Consensus
dr. Maxim Orlovsky & Pandora Foundation Team
Multi-agent systems
• AI is not singular – it is multi-agent:

autonomous vehicles, IoT, mobiles, etc.
• At some point in future(?)

agents will become interacting.
• Multiple V2V, V2E, E2E protocols will evolve,

which will face byzantine faults

and multiple security issues
Imagefromsitewww.artlebedev.ru
Distributed Systems &

Byzantine Fault Tolerance
• The Byzantine General’s Problem was first proposed by Lamport,

Shostak, and Pease in 1982.
• A problem of making a reliable system from unreliable parts

(Ethan Buchman, Cosmos Network)
• From 1982 to 1999, no one had invented a solution for the problem
• Non-blockchain based solutions (Paxos, Raft, DLS, BPFT) had limited

fault-tolerance and were used in trusted setups (space communications
etc).
• Bitcoin opened a way to enhance byzantine tolerance due to economical
stimuli and game theory. It gave rise to other blockchain technologies.
Blockchain:

Distributed Ledger Technology (DLT)
• Trusted Persistent Data Records
• Inside Trustless Environments
• Without Central Governance
• Secured By Economic Incentives
Blockchain Key Parameters
• Security: (sometimes called “immutability“ or “persistence“):

past history does not change. 

Leads to audibility and accountability
• Liveness: availability to post new transactions

Liveness intrinsically means censorship-resistance
• Blockchain-based distributed systems achieve these qualities
by following the same consensus protocol
• Non-faulty behavior is supported by 

economic incentives (like mining rewards)
Blockchain
• Replicative state machine 

(type of distributed database)
+ asymmetric cryptography (accountability)
+ ordered immutable set

of historical state changes (audibility)
Consensus Protocol
• Required for public blockchains 

(”private/federated blockchain” is an oxymoron)
• Adds rules by which distributed parts
of replicative state machine reaches consensus
on the current state
• Achieves that by adding economic incentives
for participants following the protocol utilizing
specially-designed game theory settings
(cryptocurrency)
Blockchain + Consensus Results:
• Solution to byzantine generals problems

(no other way to solve it)
• i.e. appearance of trust in trustless
environments (due to accountability and
audibility)
Technology
• Distributed database (ledger)
• Distributed computations
(state changes), 

Turing-complete OR incomplete
• Peer-to-peer mesh network
• Cryptographically secured
Economics
• Multiagent economy
• Game theory
• Free open market
• Non-state decentralized
economies linked to particular
types of resources
or businesses
BLOCKCHAIN
Block
data Block = Block
{ header :: BlockHeader
, blockSize :: Int
, transactionCount :: Int
, transactions :: [Transaction]
}
data BlockHeader = BlockHeader
{ version :: Int

, prevBlockHeader :: Hash32
, merkleRoot :: Hash32
, timestamp :: Int32
, ...
}
Hash of Previous
Block Header
Merkle Root
Block 1
Transactions
Block 1
Header
Hash of Previous
Block Header
Merkle Root
Block 1
Transactions
Block 2
Header
Smart Contracts
• Proposed by Nick Szabo much before blockchain
• Can automate multi-agent AI-based systems
• Used since bitcoin in most of blockchains
to automate state transition changes according
to some rules
• Turing-incomplete & Turing-complete
Turing-incomplete
• Bitcoin (multisigs, lightning, sidechains)
and bitcoin clones
• Limited functionality

(but robust and easily formally verified)
Bitcoin Smart Contract
<expiry time>
OP_CHECKLOCKTIMEVERIFY
OP_DROP
OP_DUP
OP_HASH160
<pubKeyHash>
OP_EQUALVERIFY
OP_CHECKSIG
<sig>
<pubKey>
Turing-complete
• Ethereum & Ethereum Classic, Rootstock 

(all EVM-based)
• Highly-vulnerable
• ”Single-threaded”, unscalable:

all computations are repeated by each node
EVM Smart Contract on Solidity
library StateMachineLib {
struct StateMachine {
bool initialized;
uint8 currentState;
mapping(uint8 => uint8[]) transitionTable;
}
function transitionToState(StateMachine storage _machine, uint8 _newState) internal {
// Checking if the state transition is allowed
bool transitionAllowed = false;
uint8[] storage allowedStates = _machine.transitionTable[uint8(_machine.currentState)];
for (uint no = 0; no < allowedStates.length; no++) {
if (allowedStates[no] == _newState) {
transitionAllowed = true;
}
}
require(transitionAllowed == true);
}
...
Problems
Why blockchains didn’t conquer the world
Consensus
How to define who has the right to sign

a new block of state changes (transactions)?
Randomness
• Physical process:

PoW – energy-consuming
• On-chain randomness:

early PoS – predictable
• Off-chain randomness:

PoS with oracles – centralised
Agreement
• BFT-algorithms (since 1980):

PoS – vulnerable
• Delegation/voting:

dPoS – centralised
• Most have problems with finality
(probabilistic at most)
Smart Contracts
• ‘Single-threaded’ global computing 

(scalability)
• Consensus depends on contract correctness
• Value depends on contract correctness
• Rich contracts can’t be truly formally verified
Non-solutions:

objectives are substituted by means
• Blockchain is neither self-aim nor any kind of distributed
technology
• “Blockchain” without cryptocurrency:

no economical incentives layer ->

no protection from not following consensus
• Private/Federative “blockchain”:

just a distributed DB with unnecessary computational burden.

no true liveness and security
• Such “blockchains” are just a word

for PR/marketing/funding purposes
Prometheus: Hybrid Consensus for AI
• Smart contracts extended to generic computing
(suitable for AI models) that can run in parallel
in trustless decentralised network
• Computing results are proved by specially-designed
PoW consensus
• Value transferred by independent PoS-consensus
• Randomness is created by computing actual models
on actual data (decentralised oracle)
Proof of Reputation
(using reputation)
Validators
Proof of Computing Work
(producing reputation)
Full Node Workers Verifiers Arbiters
Reputation level
increase
Computing work /

reward ratio increase
Prometheus Layers
Proof of Cognitive Work: 

P2P state channels

with Nash equilibrium achieved only for 

a correct AI model training/inference 

Poseidon
Proof of Reputation Blockchain: 

State Machine/Settlement Layer with 

three-tier scalability
• blockchain – Hyperion
• sidechains – Talassa
• state channels – Tethys
Poseidon
• Tasks (AI model training/inference) is split in batches.
Each batch computed in parallel by some Workers.
• Each task & batch is verified by three Verifier (using
testing set for training or repeating 10% of random
samples for inference).
• Both Workers & Verifiers have stake, 

Verifiers additionally have reputation.
• If Verifiers confirm computing then task is complete 

Worker and Verifiers get paid + Workers get mining reward
https://github.com/pandoraboxchain/pyrrha-consensus
Poseidon Arbitration
• If result is not confirmed then Worker either

can be penaltised with its stake OR it can apply
for Arbitration
• There are two-tier Arbitration procedure which provably
leads to correct result. 

Nodes found faulty are penaltised with stake/reputation
and correct nodes are rewarded with reputation & paid.
• This leads to Nash equilibrium without repeating all
computations on each node of the network.
Hyperion: replicative state machine
for settlements of Poseidon computing
• Randomness is taken from PoCW
• No Turing-complete computing
• Provable finality & security
• Liveness and censorship-resistance
• Three-tier scalability 

(blockchain, sidechains, state channels)
Imagefromsitewww.artlebedev.ru
MALICIOUS AGENT

DETECTION IN
AUTONOMOUS 

AI-DOMINATED SYSTEMS
Other Key Properties
Unique:
• Censorship resistance
• Zero governance
• Scalability
• Hybrid consensus with
PoS/PoW model
Common:
• Open markets for models,
big data and computing
power
• Economic incentives for
participants
• Zero-knowledge
What Pandora Boxchain 

will also give AI?
• Privacy & zero knowledge:

fixing “loss of privacy” problem
• Free, open markets (models, big data, computing
power): enabling faster progress and fair rewards
• Mitigating strategic risks from AI

with game theory, economics & byzantine fault
tolerance –instead of “ethics” which would not work
The Solution
Instead of:
AI regulations
Big data regulations
AI “Kill switch”
“Azimov laws”
Use:
Economic incentives
Zero knowledge
Game theory
Audibility & accountability
Blockchain
Maxim Orlovsky PhD, MD
orlovsky@pandora.foundation
/pandoraboxchain
github.com/pandoraboxchain
Special thanks to Dmitry Litvinov
for the presentation design
& Pandora Foundation Team
www.pandoraboxchain.ai
… one more thing
A bit of futurism
AI Today: Limited Progress
Engineering over science –

lack of new scientific paradigms:
• Practical progress comes from computational
resources and big data
• CNN, Deep learning, Reinforcement learning,
AGN etc represent engineering innovations
New Directions
Engineering:
• Multi-agent systems
• Distributed/
decentralised systems
• New biologically-
inspired cognitive
architectures
Science:
• Game theory
• Byzantine fault
tolerance
• Complexity science
⊂ Blockchain
Engineering:
• Multi-agent systems
• Distributed/
decentralised systems
• New biologically-
inspired cognitive
architectures
Science:
• Game theory
• Byzantine fault
tolerance
• Complexity science
happens when

selection acts

on a diverging
population
while
modern neuronets
are singular
Evolution
•species exchanges genes via viruses
•cells emit chemical and electric signals
•animals communicate via mimics and emotions
•humans — via language
Information transfer:
What about #AI?
requires

Consensus

Persistence

Liveness
Information transfer:
•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
Consensus:
each agent gets the same
sense of a signal as others
•DNA in chromosomes do not change during
the live of the most cells
•Egyptian hieroglyphs could still be
understood by researchers
Persistence:
the sense of the signal
does not diminish with time
Is perfect both for
consensus and persistence
rendering multi-agent #AI

as an evolving self-organized system
BLOCKCHAIN:

Contenu connexe

Tendances

How to Create Blockchain Products by Slice.Market CTO
How to Create Blockchain Products by Slice.Market CTOHow to Create Blockchain Products by Slice.Market CTO
How to Create Blockchain Products by Slice.Market CTOProduct School
 
Chain of a_lifetime_december2014
Chain of a_lifetime_december2014Chain of a_lifetime_december2014
Chain of a_lifetime_december2014Carlo Bertolazzi
 
Vilnius blockchain club 20170413 consensus
Vilnius blockchain club 20170413 consensusVilnius blockchain club 20170413 consensus
Vilnius blockchain club 20170413 consensusAudrius Ramoska
 
Analysing Data from Blockchains - Keynote @ SOCCA 2020
Analysing Data from Blockchains - Keynote @ SOCCA 2020Analysing Data from Blockchains - Keynote @ SOCCA 2020
Analysing Data from Blockchains - Keynote @ SOCCA 2020Ingo Weber
 
Introduction to Blockchain and Recordkeeping
Introduction to Blockchain and RecordkeepingIntroduction to Blockchain and Recordkeeping
Introduction to Blockchain and RecordkeepingPeter Van Garderen
 
Blockchain 3.0 - Decentral Applications
Blockchain 3.0 - Decentral ApplicationsBlockchain 3.0 - Decentral Applications
Blockchain 3.0 - Decentral ApplicationsBart Waeterschoot
 
Blockchain: An Introduction for Executives
Blockchain: An Introduction for ExecutivesBlockchain: An Introduction for Executives
Blockchain: An Introduction for ExecutivesScottMadden, Inc.
 
EUIPO DPM knowledge share: Blockchain and IP
EUIPO DPM knowledge share: Blockchain and IPEUIPO DPM knowledge share: Blockchain and IP
EUIPO DPM knowledge share: Blockchain and IPAudrius Ramoska
 
Blockchain Essentials for Enterprise Architects
Blockchain Essentials for Enterprise ArchitectsBlockchain Essentials for Enterprise Architects
Blockchain Essentials for Enterprise ArchitectsGokul Alex
 
BlockChain for the Banker
BlockChain for the BankerBlockChain for the Banker
BlockChain for the BankerBohdan Szymanik
 
Understanding blockchain
Understanding blockchainUnderstanding blockchain
Understanding blockchainPriyab Satoshi
 
Blockchain and Services – Exploring the Links
Blockchain and Services – Exploring the LinksBlockchain and Services – Exploring the Links
Blockchain and Services – Exploring the LinksIngo Weber
 
Blockchain - Presentacion Betabeers Galicia 10/12/2014
Blockchain - Presentacion Betabeers Galicia 10/12/2014Blockchain - Presentacion Betabeers Galicia 10/12/2014
Blockchain - Presentacion Betabeers Galicia 10/12/2014WeKCo Coworking
 
Introduction to Blockchain and Smart Contracts
Introduction to Blockchain and Smart ContractsIntroduction to Blockchain and Smart Contracts
Introduction to Blockchain and Smart ContractsSaad Zaher
 
Blockchain Technologies
Blockchain TechnologiesBlockchain Technologies
Blockchain TechnologiesAdri Jovin
 
Blockchain Application Design and Development, and the Case of Programmable M...
Blockchain Application Design and Development, and the Case of Programmable M...Blockchain Application Design and Development, and the Case of Programmable M...
Blockchain Application Design and Development, and the Case of Programmable M...Ingo Weber
 

Tendances (20)

How to Create Blockchain Products by Slice.Market CTO
How to Create Blockchain Products by Slice.Market CTOHow to Create Blockchain Products by Slice.Market CTO
How to Create Blockchain Products by Slice.Market CTO
 
Chain of a_lifetime_december2014
Chain of a_lifetime_december2014Chain of a_lifetime_december2014
Chain of a_lifetime_december2014
 
Vilnius blockchain club 20170413 consensus
Vilnius blockchain club 20170413 consensusVilnius blockchain club 20170413 consensus
Vilnius blockchain club 20170413 consensus
 
Analysing Data from Blockchains - Keynote @ SOCCA 2020
Analysing Data from Blockchains - Keynote @ SOCCA 2020Analysing Data from Blockchains - Keynote @ SOCCA 2020
Analysing Data from Blockchains - Keynote @ SOCCA 2020
 
Introduction to Blockchain and Recordkeeping
Introduction to Blockchain and RecordkeepingIntroduction to Blockchain and Recordkeeping
Introduction to Blockchain and Recordkeeping
 
Cryptocurrencies
CryptocurrenciesCryptocurrencies
Cryptocurrencies
 
Blockchain 3.0 - Decentral Applications
Blockchain 3.0 - Decentral ApplicationsBlockchain 3.0 - Decentral Applications
Blockchain 3.0 - Decentral Applications
 
Blockchain: An Introduction for Executives
Blockchain: An Introduction for ExecutivesBlockchain: An Introduction for Executives
Blockchain: An Introduction for Executives
 
Blockchains 101
Blockchains 101Blockchains 101
Blockchains 101
 
Blockchain for the Enterprise
Blockchain for the EnterpriseBlockchain for the Enterprise
Blockchain for the Enterprise
 
EUIPO DPM knowledge share: Blockchain and IP
EUIPO DPM knowledge share: Blockchain and IPEUIPO DPM knowledge share: Blockchain and IP
EUIPO DPM knowledge share: Blockchain and IP
 
Blockchain Essentials for Enterprise Architects
Blockchain Essentials for Enterprise ArchitectsBlockchain Essentials for Enterprise Architects
Blockchain Essentials for Enterprise Architects
 
BlockChain for the Banker
BlockChain for the BankerBlockChain for the Banker
BlockChain for the Banker
 
Understanding blockchain
Understanding blockchainUnderstanding blockchain
Understanding blockchain
 
Blockchain 101
Blockchain 101Blockchain 101
Blockchain 101
 
Blockchain and Services – Exploring the Links
Blockchain and Services – Exploring the LinksBlockchain and Services – Exploring the Links
Blockchain and Services – Exploring the Links
 
Blockchain - Presentacion Betabeers Galicia 10/12/2014
Blockchain - Presentacion Betabeers Galicia 10/12/2014Blockchain - Presentacion Betabeers Galicia 10/12/2014
Blockchain - Presentacion Betabeers Galicia 10/12/2014
 
Introduction to Blockchain and Smart Contracts
Introduction to Blockchain and Smart ContractsIntroduction to Blockchain and Smart Contracts
Introduction to Blockchain and Smart Contracts
 
Blockchain Technologies
Blockchain TechnologiesBlockchain Technologies
Blockchain Technologies
 
Blockchain Application Design and Development, and the Case of Programmable M...
Blockchain Application Design and Development, and the Case of Programmable M...Blockchain Application Design and Development, and the Case of Programmable M...
Blockchain Application Design and Development, and the Case of Programmable M...
 

Similaire à How AI benefits from Blockchain and Game Theory with Scalable Censorship-resistant Consensus

Why we need censorship-resistant AI
Why we need censorship-resistant AIWhy we need censorship-resistant AI
Why we need censorship-resistant AIPandora Boxchain
 
AI & Blockchain: An Introduction
AI & Blockchain: An IntroductionAI & Blockchain: An Introduction
AI & Blockchain: An IntroductionMatt Turck
 
Blockchain Future & Investments 2018 - Women in Product
Blockchain Future & Investments 2018 - Women in Product Blockchain Future & Investments 2018 - Women in Product
Blockchain Future & Investments 2018 - Women in Product Aarthi Srinivasan
 
Blockchain
BlockchainBlockchain
BlockchainSai Nath
 
On Engineering Economic Systems
On Engineering Economic SystemsOn Engineering Economic Systems
On Engineering Economic SystemsMichael Zargham
 
Nov 2 security for blockchain and analytics ulf mattsson 2020 nov 2b
Nov 2 security for blockchain and analytics   ulf mattsson 2020 nov 2bNov 2 security for blockchain and analytics   ulf mattsson 2020 nov 2b
Nov 2 security for blockchain and analytics ulf mattsson 2020 nov 2bUlf Mattsson
 
Blockchain general presentation nov 2017 v eng
Blockchain general presentation nov 2017 v engBlockchain general presentation nov 2017 v eng
Blockchain general presentation nov 2017 v engDavid Vangulick
 
Intro to blockchain - Concordium
Intro to blockchain - ConcordiumIntro to blockchain - Concordium
Intro to blockchain - ConcordiumGeorgeBenton4
 
Idea To IPO Blockchain Slides
Idea To IPO Blockchain SlidesIdea To IPO Blockchain Slides
Idea To IPO Blockchain SlidesRoger Royse
 
Blockchain with HyperLedger (Public version)
Blockchain with HyperLedger (Public version)Blockchain with HyperLedger (Public version)
Blockchain with HyperLedger (Public version)Benjamin Fuentes
 
Blockchain mechanism for IET Mauritius Network
Blockchain mechanism for IET Mauritius NetworkBlockchain mechanism for IET Mauritius Network
Blockchain mechanism for IET Mauritius Networkcharles okaformbah
 
Aleksandr Shevchenko ITEM 2018
Aleksandr Shevchenko ITEM 2018Aleksandr Shevchenko ITEM 2018
Aleksandr Shevchenko ITEM 2018ITEM
 
Blockchain-Enabled Economic Systems (ICCS 2018 Lecture)
Blockchain-Enabled Economic Systems (ICCS 2018 Lecture)Blockchain-Enabled Economic Systems (ICCS 2018 Lecture)
Blockchain-Enabled Economic Systems (ICCS 2018 Lecture)Michael Zargham
 

Similaire à How AI benefits from Blockchain and Game Theory with Scalable Censorship-resistant Consensus (20)

Why we need censorship-resistant AI
Why we need censorship-resistant AIWhy we need censorship-resistant AI
Why we need censorship-resistant AI
 
AI & Blockchain: An Introduction
AI & Blockchain: An IntroductionAI & Blockchain: An Introduction
AI & Blockchain: An Introduction
 
BCT.pptx
BCT.pptxBCT.pptx
BCT.pptx
 
Intro to Blockchain
Intro to Blockchain Intro to Blockchain
Intro to Blockchain
 
Blockchain Future & Investments 2018 - Women in Product
Blockchain Future & Investments 2018 - Women in Product Blockchain Future & Investments 2018 - Women in Product
Blockchain Future & Investments 2018 - Women in Product
 
Blockchain
BlockchainBlockchain
Blockchain
 
On Engineering Economic Systems
On Engineering Economic SystemsOn Engineering Economic Systems
On Engineering Economic Systems
 
Nov 2 security for blockchain and analytics ulf mattsson 2020 nov 2b
Nov 2 security for blockchain and analytics   ulf mattsson 2020 nov 2bNov 2 security for blockchain and analytics   ulf mattsson 2020 nov 2b
Nov 2 security for blockchain and analytics ulf mattsson 2020 nov 2b
 
The Promise of BlockChain
The Promise of BlockChainThe Promise of BlockChain
The Promise of BlockChain
 
Blockchain general presentation nov 2017 v eng
Blockchain general presentation nov 2017 v engBlockchain general presentation nov 2017 v eng
Blockchain general presentation nov 2017 v eng
 
Intro to blockchain - Concordium
Intro to blockchain - ConcordiumIntro to blockchain - Concordium
Intro to blockchain - Concordium
 
Blockchain Fundamentals
Blockchain FundamentalsBlockchain Fundamentals
Blockchain Fundamentals
 
Idea To IPO Blockchain Slides
Idea To IPO Blockchain SlidesIdea To IPO Blockchain Slides
Idea To IPO Blockchain Slides
 
Blockchain with HyperLedger (Public version)
Blockchain with HyperLedger (Public version)Blockchain with HyperLedger (Public version)
Blockchain with HyperLedger (Public version)
 
Blockchain
BlockchainBlockchain
Blockchain
 
Blockchain in a nutshell
Blockchain in a nutshellBlockchain in a nutshell
Blockchain in a nutshell
 
Blockchain mechanism for IET Mauritius Network
Blockchain mechanism for IET Mauritius NetworkBlockchain mechanism for IET Mauritius Network
Blockchain mechanism for IET Mauritius Network
 
Aleksandr Shevchenko ITEM 2018
Aleksandr Shevchenko ITEM 2018Aleksandr Shevchenko ITEM 2018
Aleksandr Shevchenko ITEM 2018
 
Blockchain-Enabled Economic Systems (ICCS 2018 Lecture)
Blockchain-Enabled Economic Systems (ICCS 2018 Lecture)Blockchain-Enabled Economic Systems (ICCS 2018 Lecture)
Blockchain-Enabled Economic Systems (ICCS 2018 Lecture)
 
blockchain-161025100639.pptx
blockchain-161025100639.pptxblockchain-161025100639.pptx
blockchain-161025100639.pptx
 

Plus de Pandora Boxchain

How to design a decentralisation: is blockchain & token needed or just a buzz...
How to design a decentralisation: is blockchain & token needed or just a buzz...How to design a decentralisation: is blockchain & token needed or just a buzz...
How to design a decentralisation: is blockchain & token needed or just a buzz...Pandora Boxchain
 
Tradeoffs in PoS Consensus Protocols
Tradeoffs in PoS Consensus ProtocolsTradeoffs in PoS Consensus Protocols
Tradeoffs in PoS Consensus ProtocolsPandora Boxchain
 
Proof of Computing Work Protocol by Pandora Boxchain
Proof of Computing Work Protocol by Pandora BoxchainProof of Computing Work Protocol by Pandora Boxchain
Proof of Computing Work Protocol by Pandora BoxchainPandora Boxchain
 
World decentralized AI on blockchain: simple explanation of Pandora Boxchain ...
World decentralized AI on blockchain: simple explanation of Pandora Boxchain ...World decentralized AI on blockchain: simple explanation of Pandora Boxchain ...
World decentralized AI on blockchain: simple explanation of Pandora Boxchain ...Pandora Boxchain
 

Plus de Pandora Boxchain (6)

The #FreeAI Manifesto
The #FreeAI ManifestoThe #FreeAI Manifesto
The #FreeAI Manifesto
 
How to design a decentralisation: is blockchain & token needed or just a buzz...
How to design a decentralisation: is blockchain & token needed or just a buzz...How to design a decentralisation: is blockchain & token needed or just a buzz...
How to design a decentralisation: is blockchain & token needed or just a buzz...
 
Malleability and SegWit
Malleability and SegWitMalleability and SegWit
Malleability and SegWit
 
Tradeoffs in PoS Consensus Protocols
Tradeoffs in PoS Consensus ProtocolsTradeoffs in PoS Consensus Protocols
Tradeoffs in PoS Consensus Protocols
 
Proof of Computing Work Protocol by Pandora Boxchain
Proof of Computing Work Protocol by Pandora BoxchainProof of Computing Work Protocol by Pandora Boxchain
Proof of Computing Work Protocol by Pandora Boxchain
 
World decentralized AI on blockchain: simple explanation of Pandora Boxchain ...
World decentralized AI on blockchain: simple explanation of Pandora Boxchain ...World decentralized AI on blockchain: simple explanation of Pandora Boxchain ...
World decentralized AI on blockchain: simple explanation of Pandora Boxchain ...
 

Dernier

Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Allon Mureinik
 
How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonetsnaman860154
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slidevu2urc
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsMaria Levchenko
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure servicePooja Nehwal
 
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...shyamraj55
 
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 3652toLead Limited
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonAnna Loughnan Colquhoun
 
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | DelhiFULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhisoniya singh
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking MenDelhi Call girls
 
GenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationGenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationMichael W. Hawkins
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitecturePixlogix Infotech
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024Rafal Los
 
Breaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountBreaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountPuma Security, LLC
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024The Digital Insurer
 
Google AI Hackathon: LLM based Evaluator for RAG
Google AI Hackathon: LLM based Evaluator for RAGGoogle AI Hackathon: LLM based Evaluator for RAG
Google AI Hackathon: LLM based Evaluator for RAGSujit Pal
 
08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking MenDelhi Call girls
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxMalak Abu Hammad
 
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...HostedbyConfluent
 
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfThe Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfEnterprise Knowledge
 

Dernier (20)

Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)
 
How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonets
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slide
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed texts
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
 
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
 
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt Robison
 
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | DelhiFULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
 
GenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationGenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day Presentation
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC Architecture
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024
 
Breaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountBreaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path Mount
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024
 
Google AI Hackathon: LLM based Evaluator for RAG
Google AI Hackathon: LLM based Evaluator for RAGGoogle AI Hackathon: LLM based Evaluator for RAG
Google AI Hackathon: LLM based Evaluator for RAG
 
08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptx
 
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
 
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfThe Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
 

How AI benefits from Blockchain and Game Theory with Scalable Censorship-resistant Consensus

  • 1. How AI benefits from Blockchain and Game Theory with 
 Scalable Censorship-resistant Consensus dr. Maxim Orlovsky & Pandora Foundation Team
  • 2. Multi-agent systems • AI is not singular – it is multi-agent:
 autonomous vehicles, IoT, mobiles, etc. • At some point in future(?)
 agents will become interacting. • Multiple V2V, V2E, E2E protocols will evolve,
 which will face byzantine faults
 and multiple security issues
  • 4. Distributed Systems &
 Byzantine Fault Tolerance • The Byzantine General’s Problem was first proposed by Lamport,
 Shostak, and Pease in 1982. • A problem of making a reliable system from unreliable parts
 (Ethan Buchman, Cosmos Network) • From 1982 to 1999, no one had invented a solution for the problem • Non-blockchain based solutions (Paxos, Raft, DLS, BPFT) had limited
 fault-tolerance and were used in trusted setups (space communications etc). • Bitcoin opened a way to enhance byzantine tolerance due to economical stimuli and game theory. It gave rise to other blockchain technologies.
  • 5. Blockchain:
 Distributed Ledger Technology (DLT) • Trusted Persistent Data Records • Inside Trustless Environments • Without Central Governance • Secured By Economic Incentives
  • 6. Blockchain Key Parameters • Security: (sometimes called “immutability“ or “persistence“):
 past history does not change. 
 Leads to audibility and accountability • Liveness: availability to post new transactions
 Liveness intrinsically means censorship-resistance • Blockchain-based distributed systems achieve these qualities by following the same consensus protocol • Non-faulty behavior is supported by 
 economic incentives (like mining rewards)
  • 7. Blockchain • Replicative state machine 
 (type of distributed database) + asymmetric cryptography (accountability) + ordered immutable set
 of historical state changes (audibility)
  • 8. Consensus Protocol • Required for public blockchains 
 (”private/federated blockchain” is an oxymoron) • Adds rules by which distributed parts of replicative state machine reaches consensus on the current state • Achieves that by adding economic incentives for participants following the protocol utilizing specially-designed game theory settings (cryptocurrency)
  • 9. Blockchain + Consensus Results: • Solution to byzantine generals problems
 (no other way to solve it) • i.e. appearance of trust in trustless environments (due to accountability and audibility)
  • 10. Technology • Distributed database (ledger) • Distributed computations (state changes), 
 Turing-complete OR incomplete • Peer-to-peer mesh network • Cryptographically secured Economics • Multiagent economy • Game theory • Free open market • Non-state decentralized economies linked to particular types of resources or businesses BLOCKCHAIN
  • 11. Block data Block = Block { header :: BlockHeader , blockSize :: Int , transactionCount :: Int , transactions :: [Transaction] } data BlockHeader = BlockHeader { version :: Int
 , prevBlockHeader :: Hash32 , merkleRoot :: Hash32 , timestamp :: Int32 , ... } Hash of Previous Block Header Merkle Root Block 1 Transactions Block 1 Header Hash of Previous Block Header Merkle Root Block 1 Transactions Block 2 Header
  • 12. Smart Contracts • Proposed by Nick Szabo much before blockchain • Can automate multi-agent AI-based systems • Used since bitcoin in most of blockchains to automate state transition changes according to some rules • Turing-incomplete & Turing-complete
  • 13. Turing-incomplete • Bitcoin (multisigs, lightning, sidechains) and bitcoin clones • Limited functionality
 (but robust and easily formally verified)
  • 14. Bitcoin Smart Contract <expiry time> OP_CHECKLOCKTIMEVERIFY OP_DROP OP_DUP OP_HASH160 <pubKeyHash> OP_EQUALVERIFY OP_CHECKSIG <sig> <pubKey>
  • 15. Turing-complete • Ethereum & Ethereum Classic, Rootstock 
 (all EVM-based) • Highly-vulnerable • ”Single-threaded”, unscalable:
 all computations are repeated by each node
  • 16. EVM Smart Contract on Solidity library StateMachineLib { struct StateMachine { bool initialized; uint8 currentState; mapping(uint8 => uint8[]) transitionTable; } function transitionToState(StateMachine storage _machine, uint8 _newState) internal { // Checking if the state transition is allowed bool transitionAllowed = false; uint8[] storage allowedStates = _machine.transitionTable[uint8(_machine.currentState)]; for (uint no = 0; no < allowedStates.length; no++) { if (allowedStates[no] == _newState) { transitionAllowed = true; } } require(transitionAllowed == true); } ...
  • 18. Consensus How to define who has the right to sign
 a new block of state changes (transactions)? Randomness • Physical process:
 PoW – energy-consuming • On-chain randomness:
 early PoS – predictable • Off-chain randomness:
 PoS with oracles – centralised Agreement • BFT-algorithms (since 1980):
 PoS – vulnerable • Delegation/voting:
 dPoS – centralised • Most have problems with finality (probabilistic at most)
  • 19. Smart Contracts • ‘Single-threaded’ global computing 
 (scalability) • Consensus depends on contract correctness • Value depends on contract correctness • Rich contracts can’t be truly formally verified
  • 20. Non-solutions:
 objectives are substituted by means • Blockchain is neither self-aim nor any kind of distributed technology • “Blockchain” without cryptocurrency:
 no economical incentives layer ->
 no protection from not following consensus • Private/Federative “blockchain”:
 just a distributed DB with unnecessary computational burden.
 no true liveness and security • Such “blockchains” are just a word
 for PR/marketing/funding purposes
  • 21.
  • 22. Prometheus: Hybrid Consensus for AI • Smart contracts extended to generic computing (suitable for AI models) that can run in parallel in trustless decentralised network • Computing results are proved by specially-designed PoW consensus • Value transferred by independent PoS-consensus • Randomness is created by computing actual models on actual data (decentralised oracle)
  • 23. Proof of Reputation (using reputation) Validators Proof of Computing Work (producing reputation) Full Node Workers Verifiers Arbiters Reputation level increase Computing work /
 reward ratio increase
  • 24. Prometheus Layers Proof of Cognitive Work: 
 P2P state channels
 with Nash equilibrium achieved only for 
 a correct AI model training/inference 
 Poseidon Proof of Reputation Blockchain: 
 State Machine/Settlement Layer with 
 three-tier scalability • blockchain – Hyperion • sidechains – Talassa • state channels – Tethys
  • 25. Poseidon • Tasks (AI model training/inference) is split in batches. Each batch computed in parallel by some Workers. • Each task & batch is verified by three Verifier (using testing set for training or repeating 10% of random samples for inference). • Both Workers & Verifiers have stake, 
 Verifiers additionally have reputation. • If Verifiers confirm computing then task is complete 
 Worker and Verifiers get paid + Workers get mining reward https://github.com/pandoraboxchain/pyrrha-consensus
  • 26. Poseidon Arbitration • If result is not confirmed then Worker either
 can be penaltised with its stake OR it can apply for Arbitration • There are two-tier Arbitration procedure which provably leads to correct result. 
 Nodes found faulty are penaltised with stake/reputation and correct nodes are rewarded with reputation & paid. • This leads to Nash equilibrium without repeating all computations on each node of the network.
  • 27. Hyperion: replicative state machine for settlements of Poseidon computing • Randomness is taken from PoCW • No Turing-complete computing • Provable finality & security • Liveness and censorship-resistance • Three-tier scalability 
 (blockchain, sidechains, state channels)
  • 29. MALICIOUS AGENT
 DETECTION IN AUTONOMOUS 
 AI-DOMINATED SYSTEMS
  • 30. Other Key Properties Unique: • Censorship resistance • Zero governance • Scalability • Hybrid consensus with PoS/PoW model Common: • Open markets for models, big data and computing power • Economic incentives for participants • Zero-knowledge
  • 31. What Pandora Boxchain 
 will also give AI? • Privacy & zero knowledge:
 fixing “loss of privacy” problem • Free, open markets (models, big data, computing power): enabling faster progress and fair rewards • Mitigating strategic risks from AI
 with game theory, economics & byzantine fault tolerance –instead of “ethics” which would not work
  • 32. The Solution Instead of: AI regulations Big data regulations AI “Kill switch” “Azimov laws” Use: Economic incentives Zero knowledge Game theory Audibility & accountability Blockchain
  • 33. Maxim Orlovsky PhD, MD orlovsky@pandora.foundation /pandoraboxchain github.com/pandoraboxchain Special thanks to Dmitry Litvinov for the presentation design & Pandora Foundation Team www.pandoraboxchain.ai
  • 34. … one more thing
  • 35. A bit of futurism
  • 36. AI Today: Limited Progress Engineering over science –
 lack of new scientific paradigms: • Practical progress comes from computational resources and big data • CNN, Deep learning, Reinforcement learning, AGN etc represent engineering innovations
  • 37. New Directions Engineering: • Multi-agent systems • Distributed/ decentralised systems • New biologically- inspired cognitive architectures Science: • Game theory • Byzantine fault tolerance • Complexity science
  • 38. ⊂ Blockchain Engineering: • Multi-agent systems • Distributed/ decentralised systems • New biologically- inspired cognitive architectures Science: • Game theory • Byzantine fault tolerance • Complexity science
  • 39. happens when
 selection acts
 on a diverging population while modern neuronets are singular Evolution
  • 40. •species exchanges genes via viruses •cells emit chemical and electric signals •animals communicate via mimics and emotions •humans — via language Information transfer: What about #AI?
  • 42. •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 Consensus: each agent gets the same sense of a signal as others
  • 43. •DNA in chromosomes do not change during the live of the most cells •Egyptian hieroglyphs could still be understood by researchers Persistence: the sense of the signal does not diminish with time
  • 44. Is perfect both for consensus and persistence rendering multi-agent #AI
 as an evolving self-organized system BLOCKCHAIN: