Decentralized artificial intelligence infrastructure project GaiaNet Inc. today announced it has raised $10 million in seed funding to provide access to edge nodes where developers can deploy artificial intelligence agents and other generative AI tools.

The company’s funding round attracted key advisors bringing additional resources including Generative Ventures’ Lex Sokolin, Republic Capital’s Brian Johnson, 7RIDGE’s Shawn Ng, Kishore Bhatia, EVM Capital, Mantle EcoFund and ByteTrade Lab.

GaiaNet operates a distributed AI infrastructure network of individually run “nodes” that allows it to decentralize model inferencing for AI applications. Instead of using centralized servers for deploying AI models, users set up individual machines to host individual models that act as individual agents customized for particular expertise, such as a teaching assistant, a business expert, a financial analyst or a crypto trader.

Similar GaiaNet nodes can be organized into domains, providing stable services under load when communities ask questions and request answers. They are accessed via an application programming interface that takes the request and then has the prompt directed to the appropriate model to the node that is ready to handle it.

“Our mission with GaiaNet is to redefine the way humans and AI engage and to enable sovereign infrastructure for interacting with AI,” said Chief Executive Matt Wright. “With the recent investment, we’re able to take a step further in our commitment to redefining the boundaries of AI, making it more accessible, data bias-resistant, collaborative and privacy-centric.”

Vertical focus

In the current phase of the project, the company is targeting outstanding issues in various industries, including higher education specifically, addressing the shortage of teaching assistants in courses with higher enrollment. By allowing the creation of personalized chatbots with accurate subject knowledge for university work that can work properly under load, GaiaNet can work as a tutor and study buddy for students in university classes, the company said.

The company partnered with the University of California at Berkeley earlier this year in an initiative to introduce distributed AI teaching assistants into its courses into more computer science and engineering courses. GaiaNet said that it believes that integrating this decentralized technology into teaching curriculums and giving students access provides a practical demonstration of the technology’s efficacy in comparison to centralized AI delivery services.

The underlying technology uses Ethereum blockchain node-based technology as part of the GaiaNet protocol, allowing it to use the decentralized nature of blockchain technology to build its networking capabilities. However, users will not see any of this in operation since it’s hidden under the surface and only node operators will interact directly with it.

GaiaNet supports numerous different AI models that can be run on its decentralized infrastructure, including more than 200 different models from HuggingFace, such as Meta Platform Inc.’s Llama 2, Google LLC’s Gemma and Mistral AI’s Mixtral family of models.

Image: geralt/Pixabay