A decentralized & distributed deep-learning training protocol
Raven protocol is a distributed and decentralized protocol for training DNN (Deep Neural Networks), through which users on the Internet can contribute to training the AI system using their desktop, laptop or smartphone and receive an incentive in return. Thanks to the Raven protocol, the cost of acquiring certain powerful processors and graphics processors for training DNN becomes minimal. Since Raven allows you to allocate resources from the simple computing power of the devices of individual participants. The concept of sharing computing power for training significantly reduces costs. In turn, depositors receive compensation / reward using Raven tokens (RAV).
The founder, Sherman Lee, is working on decentralized AI training. He is a partner of @Zeroth_AI - a company that finances AI / blockchain projects.
Now the training processes of algorithms are in great demand on the market and are used in various spheres of life. It is the processing of video, voice recognition, improving the quality of photos. So far, only giant companies can afford mass use. Therefore, we see a certain monopoly, but, perhaps, with the arrival of the blockchain in this sphere, the game will change.
At this stage, of course, there is little information: no WP, no funds are indicated. There is no hype and much discussion about the project, but the idea is certainly interesting. At the moment, the Raven team has people who really understand the AI and Blockchain sphere. The project is being developed for several years and the team could hold ICO for a long time, but they did not do it. They focus on creating MVP. And it seems that Raven communicates quite closely with the Binance exchange team.
Available for token sale