In a way, it is similar to our approach at Q Blocks. Decentralized GPU computing for ML.