Ezra Palmers

Statistics and probability have always been the ground I stand on - they came naturally in a way little else did. Machine learning was an easy step from there: it is the applied face of probability, joined to code and computation, and it is where I have put my energy for the past two years.

What interests me is the problem, not the plumbing. Modern coding tools have removed most of the friction of writing software, which means the time I spend is on understanding a system or a domain properly and reasoning about the right model for it - the part I was always better at anyway.

I am looking for what comes next to be genuinely hard: machine learning or deep learning work with real mentorship, closer to research than to reporting.