This is impressive and amazing. I can already see a press release not too far down the road: "Our new model HoSapiens can do everything humans can do, but better. It has been specifically designed to deprecate humanity. We are working with red teamers — domain experts in areas like union busting, corporate law and counterinsurgency, plus our habitual bias, misinformation, and hateful content against AI orange team— who will be adversarially testing the model.