New blockchain consensus uses AI fuzzy logic for fairer validators

A team of researchers led by Javier Andreu-Perez has published a framework that rethinks blockchain consensus by incorporating reputation awareness. Traditional consensus mechanisms like Proof-of-Work or Proof-of-Stake often require massive computational power or high stakes, leading to centralization. The new approach uses intuitionistic fuzzy sets (IFSs) to capture the inherent uncertainty in a validator's reputation. Unlike simple scores, IFSs explicitly model both membership and non-membership degrees, reflecting the lack of precise knowledge.

To track reputation over time, the framework employs uninorm aggregation operations that give balanced weight to both positive and negative feedback. This allows validators who made mistakes to later redeem themselves by performing well, fostering a more forgiving and equitable system. The algorithm maintains linear computational complexity and adds no extra communication overhead beyond the underlying consensus protocol. Experimental results show improved performance and fairness, making it a promising step toward decentralized and inclusive blockchain networks.