SquiggleAI

active

An AI-assisted tool hosted on Squiggle Hub that uses large language models to generate, fix, and improve Squiggle code from natural-language prompts.

Endorsements support Quantified Uncertainty Research Institute.

An AI-assisted tool hosted on Squiggle Hub that uses large language models to generate, fix, and improve Squiggle code from natural-language prompts.

Endorsements support Quantified Uncertainty Research Institute.

People– no linked people

Funding Details

Start Date: Oct 3, 2024
End Date: -
Expected Duration: -
Funding Raised to Date: -
Annual Budget: -
Monthly Burn Rate: -
Current Runway: -
Funding Goal: -
Funding Stage: -
Fiscal Sponsor: -

Project Details

SquiggleAI is a QURI-built system that orchestrates large language model calls to automatically produce probabilistic models in the Squiggle language, iteratively refining code until it runs and captures the user’s intent. It is integrated into Squiggle Hub and is intended both as a practical assistant for building cost-effectiveness and forecasting models and as an experiment in AI-assisted epistemic reasoning.

Theory of Change

SquiggleAI is intended to remove the programming barrier to probabilistic modeling by letting domain experts describe problems in natural language and automatically generating Squiggle models, thereby enabling more people to build and iterate on quantitative decision analyses.

Grants Received– no grants recorded

Discussion

No comments yet. Be the first to share your thoughts.