How to run neural networks (AI) in a decentralised network? How to solve scalability problem for large-scale and high-load computing in blockchain apps? The presentation describes Proof of Computing Work protocol that has Nash equilibrium for provably correct computations in a trustless environment without repeating them on each of the network nodes.
6. What we’ve done so far
Proof of Computing Work (PoCW)
Protocol that allows large-scale computing (like AI)
in a trustless peer network without
repeating computations on each of the nodes
– this solves main blockchain & decentralised networks
scalability problem
Consensus protocol (Prometheus)
Based on top of PoCW, solves main PoW and PoS problems
7. Proof of Cognitive Work
Scalability protocol reducing computation redundancy from
O(N) to O(1), where N is the number of nodes in
decentralised network
● Scientifically proved with game
theory
● Formally specified in UML
● Implemented in Ethereum smart
contracts
State of PoCW:
● Implemented python node able
to perform training and
inference on Keras neural
networks
● Testnet web interface
● PoCW web explorer
8. PoCW Basics
● ML computation run only once by a randomly selected worker node
of the network. Node puts a stake on the correctness of the
computing.
● 10% of the computation is verified once by a higher-staked node
(Verifier), randomly selected also from all high-staked nodes
● In case of inconsistency (negative verdicts of computation
correctness) the stake of the original worker is slashed
● Worker can put an appeal (by increasing stake).
Two round arbitration (closed and public) is run in this case
(details follow)
12. Input data split on batches
Assigned worker node to process each batch
Batched results of data processing
Hashes of results
Verificators re-computing 10% of batches
Verdicts
Verification: Verdicts
Model
Model
19. Full node
Proof of Computing Work
(producing reputation)
Workers Verifiers Arbiters
Reputation
level
increase
Computing
work/reward
ratio increase
20. Where are we with release
● Science: formal verification, academic papers prepared
● Engineering: tech specifications in UML
https://github.com/pandoraboxchain/pyrrha-techspecs
● Code: smart contracts + full node code
https://github.com/pandoraboxchain/pyrrha-consensus
https://github.com/pandoraboxchain/pyrrha-pynode
● Testnet in Ethereum network + Web UI to testnet
(web client, explorer)
http://client.pandora.network/
http://pandora.network/
21. Formal verification of
PoCW consensus with game
theory:
probabilistic prove for
computation correctness in
trustless environment with O(1)
for a network with N nodes
27. Core Pandora Boxchain Differentiators
● Censorship resistance
● Decentralised AI
(and generic large-scale
parallel and sequential)
computing
● Generic setup with economic
incentivisation of
5 types of actors
For this aims it …
● … provides own protocol for
proving correctness of
computation in trustless
network without repeating
all computing on each of
the nodes
● … provides its own
consensus protocol
28. Ask your question on Telegram:
https://t.me/pandoraboxchain_chat
Follow on Twitter:
https://twitter.com/pandoraboxchain