Squiggle is a special-purpose programming language for probabilistic estimation. Think: “Guesstimate as a programming language.” Squiggle is free and open-source.
Read the overview.
Overarching goals
Allow non-experts to see the “sum” of views on a question
Allow non-experts
Allow expression of estimated distributions and functions easily [1]
Squiggle is not
For intensive calcution
Models vs estimation functions
A key notion is the difference between models and estimation functions
Models are high detail and expensive to run. They make implications about causality
Explanation functions used cached date and are cheap to run
The world is a model, a photograph or video game simulator is an estimation function
The world is made of complex models, but we want answers to simpler questions—what is the most pressing thing to work on? How do we reduce risk? What am I going to have for lunch?
How do we allow people to read some set of estimation functions on different problems?
If you want to know what an estimate of AI risk is, how do you see people’s simple models (estimation functions) of this?
Some questions
Why did you do squiggle when it doesn’t seem to deal with this estimation functions problem
Why not build a python package?