HN Showcase
Back to browse
Noise Lang, JIT stochastic programing language

Noise Lang, JIT stochastic programing language

by manucorporat·Jun 27, 2026·1 point·0 comments
Visit ProjectView on HN

AI Analysis

●●●BangerBig BrainWizardryZero to One

Nine-year language where every variable is a probability distribution, not a value.

Strengths
  • •Operators lift over distributions automatically—no manual Monte-Carlo boilerplate.
  • •JIT compilation makes stochastic programs actually fast, not just academic.
  • •Interactive playground shows distributions unfolding as you scroll.
Weaknesses
  • •Niche audience—probabilistic programming is already served by Stan, PyMC.
  • •Nine years but only now 'fully working' with Claude's help raises questions.
Target Audience

Data scientists, researchers, developers working with uncertainty modeling

Similar To

Stan · PyMC · Church

Post Description

9 years ago, I started this side project to learn Rust and create a programming language where every variable represent a probability distribution. Found it very useful to describe noise in telecommunications, and have some fun with interesting problems. After early abandoning it, used Claude to get it into a fully working state. Hope you like it.

Similar Projects