Future Industrial Internet-of-Things in the upcoming 6G era is expected to deploy artificial intelligence (AI) and digital twins (DTs) ubiquitously. As a complement to conventional AI solutions, emergent intelligence (EI) exhibits various outstanding features including robustness, protection to privacy, and scalability, which makes it competitive for 6G IIoT applications. However, despite its low computational complexity, it is challenged by its high demand of data traffic in massive deployment. In this paper, we propose to exploit the massive twinning paradigm, which 6G is envisaged to support, to reduce the data traffic in EI and therewith enhance its performance.