Poseidon obtained $ 15 million in seed funding led by A16Z Crypto to build a decentralized data layer designed for artificial intelligence training.
The AI Complete data layer based in San Francisco said that it was aimed at combating the scarcity of high -quality training data and approved in IP in AI development, according to an announcement Tuesday shared with Cintelelegraph.
“The LLM and the calculation are no longer the bottlenecks; these are high quality data that are missing,” said Sandeep Chinhali, chief scientist of Poseidon and chief of the chief AI of his incubator, history protocol.
“Poseidon provides the sets of real-structured and structured IP data that AI teams need to create systems that actually work in physical and complex environments,” he added.
In relation: Oort ai decentralized data hit the best ranks on Google Kaggle
Decentralized pipeline for AI legal training data
The Poseidon solution is based on a decentralized infrastructure to collect and distribute sets of data legally deleted for commercial use. The platform incorporates the ONCHAIN de Story license infrastructure to ensure traceability and monetization, allowing data contributors to be paid for their work while protecting developers from intellectual property risks.
The team argues that centralized data supply models cannot meet the growing demand for niche and high context data sets necessary for new generation AI models, especially in fields such as robotics and space computer science.
Chris Dixon, founder of A16Z Crypto, described the project as a step towards “a new economic base for the Internet”. He added that the model rewards creators and suppliers to “provide the various inputs whose new generation intelligent systems need”.
Poseidon works with several AI laboratories and plans to use funding to evolve its infrastructure. This includes the launch of contributor modules, software development kits and license tools for developers and data suppliers. Early access should start this summer.
In relation: Bold Technologies and Mon Aion launch $ 2.5 billion on the IA platform of the smart city
Poseidon to resolve drought of AI data
The first wave of AI foundation models prospered on many online data, but this era is over, said analysts A16z Chris Dixon and Carra Wu in a note shared with Coingraph.
They noted that easily accessible data sets, including books, websites and public recordings, have been widely used, leaving hungry AI models for fresh, high quality and legally usable information.
“The challenge is not only technical – this is a coordination problem. Thousands of contributors must work together in a distributed manner to source, label and maintain the physical data that the new generation AI needs,” wrote the duo.
https://www.youtube.com/watch?v=NSZ0CYHCIF
They added that no centralized approach can effectively orchestrate the creation and preservation of the necessary data at the level of scale and diversity required. “A decentralized approach can solve this problem,” they said.
Review: AI EYE: An increasing number of users take LSD with Chatgpt
