“Phil Trammell argues that a leading AI company could be a widely-held public entity, which would provide a safety buffer without concentrating wealth, mitigating the perceived trade-off between AI safety and commoditization.”