Based on the Passage work, Galois started getting involved with Probabilistic Programming for Advancing Machine Learning (PPAML), but they needed someone to serve as technical lead.
What were the biggest challenges in developing probabilistic programming for a new language?
When people talk about Probabilistic Programming Languages (PPLs), they usually mean a system for building and reasoning about Bayesian models.
So our system has an expressive language for reasoning about relational structure, and works in terms that make these optimizations natural for machine learning and probabilistic programming.
There seems to be an explosion in probabilistic programming on Julia with other libraries like Turing or Gen, how does Soss compare to them?