Illia Polosukhin, co-founder of NEAR Protocol, recently unveiled the master plan for its AI R&D lab, NEAR AI. The plan is divided into three phases and aims to achieve open source and user-owned general artificial intelligence (AGI). Illia Polosukhin has a deep background in AI and was one of the contributors to Google's AI project TensorFlow and co-builder of the Transformer neural network architecture, which is an important underlying technology for natural language processing and AI today.

The first phase of NEAR AI is AI Developer, the goal of which is to teach machines to program. The second phase is AI Researcher, in which AI Developer will be used to teach machines to conduct research. The last phase is to use AI researchers to advance science and move towards a universal AGI shared by all.

In addition, NEAR Protocol has also launched NEAR Tasks, a blockchain-based AI annotation platform. On this platform, the demand side (Vendor) of model training can issue task requests and upload basic data materials, while users (Taskers) can participate in task answering, text annotation, image recognition and other manual operations. After completing the task, users will be rewarded with NEAR tokens, and these manually labeled data will be used to train the corresponding AI models.

Illia Polosukhin emphasized that the development of NEAR AI will always be open source, providing software, data sets, and models to the wider community to further develop other products. To achieve this goal, they need a range of infrastructure beyond the core blockchain primitives, including peer-to-peer communication, edge data and reasoning, decentralized data storage, private computing, and more. All of this is only possible by participating in the entire NEAR ecosystem and leveraging the existing NEAR token economy$NEAR #Near格局很大