I studied Mathematical Statistics with David Siegmund and Herbert Robbins at Columbia University in the 1960s. My dissertation was in sequential analysis: how to estimate the mean of a distribution with prescribed accuracy when the variance is unknown. I had a long statistical career at IBM Research, making statistical models for all sorts of things, doing mathematical reliability and modeling of the computer. I spent some years working on speech recognition, a pattern recognition project whose results became the IBM speech recognition system Via Voice and its prodigy. One may now buy this for under $100. Eight years ago I left IBM and joined the faculty at New York University, where I work in biostatistics and mathematical biology in the Medical School. I have developed a model for estimating the rate of spontaneous mutageneis in vitro. I am currently working with a group developing a polyvalent vaccine against HIV, and in connection with this I am particularly interested in models of the immune system and the immune response.