Trent McConaghy
@trentmc0
Data, AI and Tokens:
Ocean Protocol
Audio radar
1000x more data
The Unreasonable Effectiveness of Data
Deep learning
models with >> capacity
Error 5% .. 0.01%Models with
limited capacity
Error 25% .. 5%
Another 1000x more data
Deep Learning Loves Data
The Problem
Mo’ data
(and mo’ compute)
Mo’ accuracy
Mo’ $
Here’s
your
personal
data
Eroding the silos
Silo Pool Mo’ Data
(and mo’ compute)
Mo’ accuracy
Mo’ $
Have lotsa data
(1000 enterprises)
Have lotsa AI
(1000 AI startups)
Have lotsa AI
(1000 AI startups)
Have lotsa data
(1000 enterprises)
A new data economy
Have lotsa AI
(1000 AI startups)
Have lotsa data
(1000 enterprises)
DM DM DM DM
DM DM DM DM
A new data economy
Have lotsa AI
(1000 AI startups)
Have lotsa data
(1000 enterprises)
DM DM DM DM
DM DM DM DM
Ocean
Economic incentives
Blockchain Superpower:
Get people to do stuff
By rewarding with tokens
Bitcoin goal: maximize security of network
Token rewards if: run compute to secure network
Economic Incentive for Bitcoin
Objective: Maximize security of network
• Where “security” = compute power
• Therefore, super expensive to roll back changes to the transaction log
E(Ri) α Hi * T
E() = expected
value
# tokens (BTC)
dispensed each
block
block
rewards
hash power of actor
= contribution to
“security”
Ocean goal: maximize supply of relevant data
Token rewards if: supply data, and curate it
Economic Incentive for Ocean
Objective: Maximize supply of relevant data
• This means: reward curating data + making it available
• Where “curating” = betting on data. Reward taste-making.
E(Rij) α log10(Sij) * log10(Dj) * T *Ri
Expected
reward for user
i on dataset j
Dj = proofed popularity
= # times made dataset
available
Sij = predicted popularity
= user’s curation market
stake in dataset j
# tokens
during
interval
From AI data to AI services
Motivations:
• Privacy, so compute on-premise or decentralized
• Data is heavy, so compute on-premise
• Link in emerging decentralized AI compute
Objective function: Maximize supply of relevant services
=reward curating services + proving that it was delivered
E(Rij) α log10(Sij) * log10(Dj) * T *Ri
proofed popularity
of service
predicted popularity
of service
Ocean is a network of curated services. An AI services pipeline.
Availability Consumption Privacy GovernanceProduction
commons
Inter-
Operability
Discovery
*Note: logos shown are examples and do not imply partnerships or integrations
Prototypes
Sandbox with Singapore Data Authority
What unlocking AI data
& services unlocks
Self-driving cars: fewer accidents, more mobility
>100x more data for health
care research
Erode the data silos
Erode the data silos
Erode the data silos
Conclusion
AI data is siloed.
AI services are siloed.
Let’s change the rules of the game with incentives.
Let’s democratize access to AI data & services!
Trent McConaghy
@trentmc0

Data, AI, and Tokens: Ocean Protocol

  • 1.
    Trent McConaghy @trentmc0 Data, AIand Tokens: Ocean Protocol
  • 2.
  • 3.
    1000x more data TheUnreasonable Effectiveness of Data
  • 4.
    Deep learning models with>> capacity Error 5% .. 0.01%Models with limited capacity Error 25% .. 5% Another 1000x more data Deep Learning Loves Data
  • 5.
  • 6.
    Mo’ data (and mo’compute) Mo’ accuracy Mo’ $
  • 7.
  • 9.
  • 10.
    Silo Pool Mo’Data (and mo’ compute) Mo’ accuracy Mo’ $
  • 11.
    Have lotsa data (1000enterprises) Have lotsa AI (1000 AI startups)
  • 12.
    Have lotsa AI (1000AI startups) Have lotsa data (1000 enterprises)
  • 13.
    A new dataeconomy Have lotsa AI (1000 AI startups) Have lotsa data (1000 enterprises) DM DM DM DM DM DM DM DM
  • 14.
    A new dataeconomy Have lotsa AI (1000 AI startups) Have lotsa data (1000 enterprises) DM DM DM DM DM DM DM DM Ocean
  • 15.
  • 16.
    Blockchain Superpower: Get peopleto do stuff By rewarding with tokens
  • 17.
    Bitcoin goal: maximizesecurity of network Token rewards if: run compute to secure network
  • 18.
    Economic Incentive forBitcoin Objective: Maximize security of network • Where “security” = compute power • Therefore, super expensive to roll back changes to the transaction log E(Ri) α Hi * T E() = expected value # tokens (BTC) dispensed each block block rewards hash power of actor = contribution to “security”
  • 19.
    Ocean goal: maximizesupply of relevant data Token rewards if: supply data, and curate it
  • 20.
    Economic Incentive forOcean Objective: Maximize supply of relevant data • This means: reward curating data + making it available • Where “curating” = betting on data. Reward taste-making. E(Rij) α log10(Sij) * log10(Dj) * T *Ri Expected reward for user i on dataset j Dj = proofed popularity = # times made dataset available Sij = predicted popularity = user’s curation market stake in dataset j # tokens during interval
  • 21.
    From AI datato AI services Motivations: • Privacy, so compute on-premise or decentralized • Data is heavy, so compute on-premise • Link in emerging decentralized AI compute Objective function: Maximize supply of relevant services =reward curating services + proving that it was delivered E(Rij) α log10(Sij) * log10(Dj) * T *Ri proofed popularity of service predicted popularity of service
  • 22.
    Ocean is anetwork of curated services. An AI services pipeline. Availability Consumption Privacy GovernanceProduction commons Inter- Operability Discovery *Note: logos shown are examples and do not imply partnerships or integrations
  • 23.
  • 25.
    Sandbox with SingaporeData Authority
  • 26.
    What unlocking AIdata & services unlocks
  • 27.
    Self-driving cars: feweraccidents, more mobility
  • 28.
    >100x more datafor health care research
  • 29.
  • 30.
  • 31.
  • 32.
  • 33.
    AI data issiloed. AI services are siloed. Let’s change the rules of the game with incentives. Let’s democratize access to AI data & services! Trent McConaghy @trentmc0