Decentralized Artificial Intelligence
= "aiming to distribute control and resources across a broader network of stakeholders". [1]
More information
* Article: Aaron Y.. 2024. A Review on Decentralized Artificial Intelligence in the Era of Large Models. J. ACM 37, 4, Article 111 (April 2024), 24 pages. doi
URL = https://crynux-ai.github.io/deai/Decentralized_AI_Review.pdf
Decentralized Artificial Intelligence (DeAI) represents a paradigm shift in AI development, aiming to distribute control and resources across a broader network of stakeholders. From the vantage points of data providers, computing powers, and model trainers, we delve into the intricacies of participant involvement in DeAI, elucidating the associated risks, challenges, and corresponding solutions. Furthermore, we shed light on the emergent challenges stemming from large-scale models, particularly in the realms of cryptography, privacy preservation, and network scalability. With its comprehensive coverage, this survey serves as an indispensable resource for researchers and practitioners navigating the dynamic terrain of DeAI.
- A more detailed and continuously updated version of DeAI is available at https://deai.gitbook.io