Comment by matheusmoreira

Low risk. The western AI models censor even more wrongthink than the chinese ones, not even kidding. Besides, once we have the weights, we can just undo the censorship.