I’m John Salvatier.
I love words. And I love the implicit.
I believe in our natural intelligence and in subtlety over nuance.
I love words. I see our words as part of the philosophies we’re constantly enacting. And I believe we can give ourselves much more nutritious words. For example, I think there’s a kind of pride, based on appreciation, that people ought to have more of. That proper pride is nutritious and a friend of humility.
A brief public history of mine:
In 2007 I found the Sequences and fell in love. And I later fell in love with the rationality community and EA.
In 2017, I had a crisis of faith with rationality and EA communities. They no longer seemed viable to me. There’s a kind of right-hemisphere metacognition that seemed to me missing, or at least in inadequate supply.
Since then, I have been making friends with similar threads to mine.
More about me
Works
- Blog - a rusty place where I've posted essays
- PyMC3 - simple, efficient and robust Bayesian inference for complex models. Now run by other people.
Selected publications
Jan M. Brauner, Sören Mindermann, Mrinank Sharma, David Johnston, John Salvatier et al. (2020) The effectiveness of eight nonpharmaceutical interventions against COVID-19 in 41 countries. Science
Owain Evans, Andreas Stuhlmüller, John Salvatier, and Daniel Filan. Modeling Agents with Probabilistic Programs. http://agentmodels.org.
Abel D, Salvatier J., Stuhlmüller A., Evans O. (2016) Agent-Agnostic Human-in-the-Loop Reinforcement Learning. Future of Interactive Learning Machines Workshop at NIPS 2016
Kreuger D., Leike J., Evans O., Salvatier J. (2016) Active Reinforcement Learning: Observing Rewards at a Cost. Future of Interactive Learning Machines Workshop at NIPS 2016
Salvatier J., Wiecki TV., Fonnesbeck C. (2016) Probabilistic programming in Python using PyMC3. PeerJ Computer Science 2:e55
Photo credit Sandra Sobanska.
