Full threadinterg12·Because someone needs to say it: We should all be highly skeptical of a single model's results.View on HN