kqr comments on Is “superhuman” AI forecasting BS? Some experiments on the “539″ bot from the Centre for AI Safety