Mitchell Reynolds comments on Estimating the Current and Future Number of AI Safety Researchers