A Very Short Introduction to Decentralized Artificial Intelligence (DeAI)
Today, the majority of the research & development in the artificial intelligence industry is done within large centralized organizations. These may be university research labs but mostly it happens inside the walls of mega corporations.
Just a handful of large technology companies and centralized entities have a monopoly over datasets for training, compute resources, and research into the latest breakthrough algorithms. They have no incentive to share this with anyone outside of their organizations. Small startups without massive resources are just left hanging out to dry, essentially preventing innovation.
Should these innovations be democratized for the rest of the world?
The DeAI community thinks so!
DeAI Community Mission Statement
We want to make sure that AI, as one of the biggest sources for innovation these days, is not left in the hands of a select few.
Working with AI is resource intensive. It needs tons of high-quality data, incredible compute power, and access to breakthrough algorithms. Large corporate organizations have access to these resources, startups don’t, the community doesn’t.
DeAI will enable the next 10x leap of creating value from AI by establishing a digital ecosystem around data, compute, and algorithms, and giving everyone access to those resources.
The Decentralized Artificial Intelligence community encompasses and welcomes all projects, teams, and individual contributors. With more people innovating, a greater part of its potential can be made available for everyone.
The DeAI Ecosystem
The DeAI ecosystem is being created by crypto projects along the data value chain: From connecting to data sources, aggregation and enrichment of data, to analysis for insights generation, to automation environments for AI algorithms.
Our Vision for the Future of DeAI
The new tool and services stack that DeAI projects are creating allows creating AI innovation outside mega corporations. Many projects are working together to create the full stack of services so a data scientist can access data, train her algorithms, and share her findings.
When looking for an AI model to reuse or adapt, data scientists can browse decentralized model marketplaces or order new models.
To either further develop their models, they can get data on data marketplaces. Data marketplaces might already contain interesting raw data from IOT devices or streaming data, or data scientists could get them directly from the source.
If the raw data quality is not high enough or data needs to be annotated, data scientists can use decentralized data curation networks. And to ensure security, safety, and integrity of the data, data scientists can use decentralized data security networks.
Once an AI model has been developed, data scientists can train it on specialized decentralized compute networks. A trained AI model can then be put back on decentralized model marketplaces to share it with the community.
Or it can be released into the wild by deploying it to a decentralized model runtime environment where it uses data oracles for continuous input of up to date data and do what it has been designed for.
Who we are and where you can find us
The DeAI community brings together AI/ML engineers, open source contributors, industry experts, pioneers, investors, researchers, and many more supporters globally to push the entire decentralized AI industry forward.
Everyone is welcome and everyone can join!
You can find us here:
DeAI community Discord
DeAI community Twitter
Sehound.eth (hut34 and theagora)
Sherman Lee (Raven Protocol)
Sebastian Wurst
