Talk at the World Summit AI 2019 by Diwaker Gupta. Covers intersection of Blockchain and AI, and specifically how Blockchains might offer some solutions for bringing ML capabilities to decentralized applications.
6. Most ML today is
centralized
Data moats (often user
generated)
Expertise
Resources
7. Most ML today is
centralized
Data moats (often user
generated)
Expertise
Resources
Democratize ML?
Users control data
Limited expertise
Limited resources
10. Readily available datasets
Off-the-shelf models
Commercial APIs
(e.g. image classification)
Easy, right?
Kate Crawford and Trevor Paglen, “Excavating AI:
The Politics of Training Sets for Machine Learning
(September 19, 2019) https: //excavating.ai
11. MSR: Decentralized & Collaborative AI
on Blockchain
“framework to host and train publicly available machine learning
models” — July 2019, https://github.com/Microsoft/0xDeCA10B
12. Cold Start Problem
No Data and/or
No Models
Limited infrastructure
Privacy limitations
13. Cold Start Problem
No Data and/or
No Models
Limited infrastructure
Privacy limitations
Blockchains can facilitate
effective marketplaces
15. Comprehensive Marketplace for Data & Services
An ecosystem for the data economy and associated services,
with a tokenized service layer that securely exposes
data, storage, compute and algorithms for consumption.
(source: oceanprotocol.com)
16. • Jupyter instance at https: //datascience.oceanprotocol.com
• Data Marketplace at https: //commons.oceanprotocol.com
• Participate in the Data Challenge! https: //
oceanprotocol.com/challenge/