I think this AI safety experiment is solid, it should go way further than this.
I think this AI safety experiment is solid, it should go way further than this.