Source: Rishin Sharma & Jake Brukhman, "Open Neural Networks: The Intersection of AI and Web3," avail. online
The Context.
When disillusionment surrounding the lack of high-integrity building in the crypto space sets in, I return to a problem space of high importance: AI/ML.
While much discourse is seemingly around stable diffusion, incredible chat bots, stunning images, and the like, my attention is drawn to the substrate: data, compute, (economic) coordination.
Today, I read the subject piece, which neatly identifies core infrastructure bottlenecks in the development of (ethical) AI/ML workflows:
(1) data availability and quality
(2) computational burden (see, e.g. Field Notes 45, 51, 64)
(3) incentivization needed to attract resources into an open source environment
Unlike many areas of crypto, in which participants perfunctorily wave the banner of decentralization (indeed, the term has served as a floating signifier in the industry), the domain of AI/ML has a great deal to do with the decentralization / centralization gradient, ethically and economically. (See, e.g., Field Note 52 and 71).
The Call.
Who is working on decentralized AI/ML, federated learning, and closely-knit areas (a) theoretically, practically, and everything in between (b) at any stage (ideation to advanced stage) with (c) a combined ethical + economic imperative?
How can I turbocharge your efforts?
The Details.
