DECENTRALISED AI
THE WORLD AI
AUTHOR: ANH LE
Artificial Intelligence (AI) technology in its present state is centralised, which means it is dependent on corporate infrastructure and, therefore, limited by the power, mobility and connectivity of its host network. Decentralised AI aims to eliminate these limitations by allowing AI technology to operate on decentralised infrastructure.
Problems of
existing AI solutions
- Work like isolated neurons or just 10% of the human brain
- No data sharing between existing AI solutions
- Limited data storage
- Limited data sample
- Limited output results
- AI’s intelligence is limited
- AI is only centralised, and depends on corporation’s infrastructure
A powerful new idea and approach
- Network connections as neural links
- Network shards as neurons
- Network database as brain’s storage capacity
- Decentralised Apps (or dApps) as I/O
- IoT as sensors
- Algorithms stay the same
- Newly conceived architecture
The solution
This architecture that mimicks our human brain has been designed specifically for the purpose of decentralised AI.
Time segment shards mimic neural storage
- Active Nodes
- Next Active Nodes
- Substitute Nodes
- Backup Nodes
The coordinator layer mimics neural links and the human brain’s working mechanism
Open source AI on a decentralised infrastructure is the strongest AI
Menu
68 Circular Road,
#02-01, 049422,
Singapore
© All rights reserved – Bytese 2019