Announcing Squiggle Hub

web

2023·EA Forum·forum.effectivealtruism.org/posts/rK3NgqLg3HHDzyLah/annou...

Authors

Credibility Rating

3/5

Good(3)

Good quality. Reputable source with community review or editorial standards, but less rigorous than peer-reviewed venues.

Rating inherited from publication venue: EA Forum

Squiggle Hub is developed by QURI (Quantified Uncertainty Research Institute) as part of their broader effort to improve quantitative reasoning tools for EA and AI safety risk estimation; relevant for those building cost-effectiveness or risk models.

Forum Post Details

Karma

131

Comments

Forum

eaforum

Forum Tags

Cause prioritizationForecastingSquiggleQuantified Uncertainty Research InstituteModelsSoftware engineering

Metadata

Importance: 38/100blog postnews

Summary

Squiggle Hub is a platform for sharing, discovering, and collaborating on probabilistic estimation models built with the Squiggle programming language. It enables the EA and forecasting community to publish quantitative models publicly, making uncertainty quantification and cost-effectiveness analyses more accessible and reproducible. The announcement introduces key features like model versioning, collaboration tools, and a public gallery of models.

Key Points

•Squiggle Hub is a web platform for hosting and sharing Squiggle probabilistic models, analogous to GitHub for code or HuggingFace for ML models.
•Users can publish quantitative models publicly, enabling others to view, fork, and build upon existing uncertainty and cost-effectiveness analyses.
•The platform supports versioning and collaboration, improving reproducibility and transparency of quantitative reasoning in EA contexts.
•Squiggle is a domain-specific language designed for expressing probability distributions and running Monte Carlo estimates easily.
•Aims to foster a community norm of sharing quantitative models, making forecasting and decision analysis more open and iterative.

1 FactBase fact citing this source

Entity	Property	Value	As Of
Squiggle Hub	Founded Date	2023	—

Cached Content Preview

HTTP 200Fetched May 17, 202612 KB

# Announcing Squiggle Hub
By Ozzie Gooen, Slava Matyukhn
Published: 2023-08-05
![](https://lh4.googleusercontent.com/vVceI1DYcTFbQt5oTVZsqOnlXnMvj-iHBkrejM6CZ13W57obHFvFfJjQDjeYowuiRVGmqcDoUgO4F9kz6Htf3GUk4FeKlMT1Kq8hvi4_f_9g5lzbLeBULphDFDPxLQmEkkw2y3VxLtZziGvxEuVJwnM)

[Very simple example](https://squigglehub.org/users/ozziegooen/models/costs-of-sfo-to-uk-flight-claude), for demonstration 

Overview
--------

[Squiggle Hub](https://squigglehub.org) is a platform for the creation and sharing of code written in[ Squiggle](https://www.squiggle-language.com/). As with Squiggle, Squiggle Hub is free and[ open-source](https://github.com/quantified-uncertainty/squiggle). 

As a refresher, [Squiggle](https://www.squiggle-language.com/) is a simple programming language for probabilistic estimation that runs on Javascript. It begins with the syntax of Guesstimate, but generally adds a lot more functionality. See its [launch post here](https://forum.effectivealtruism.org/posts/TfPdb2aMKzgWXFvc3/announcing-squiggle-early-access) for more information, or [the website](https://www.squiggle-language.com/) for the full documentation.

Squiggle Hub is a lot like a more powerful, but less visual, version of[ Guesstimate](https://www.getguesstimate.com/). We hope that it will eventually be much more valuable than Guesstimate is now. 

If you can use Guesstimate, you can basically use Squiggle. If you already use Guesstimate, try using the same syntax in Squiggle. It should mostly work.

All models on Squiggle Hub are public. We've produced several small ones so far, and a few friends have written some as well. We're looking forward to seeing what others make!

Looking for Squiggle examples? We’ve organized [some in the docs](https://www.squiggle-language.com/docs/Ecosystem/Gallery). The [Squiggle EA Forum Tag](https://forum.effectivealtruism.org/topics/squiggle) also has an updating list. 

### **Key Links**

*   [Squiggle Hub](https://squigglehub.org/)
*   [Squiggle Discord](https://discord.gg/nsTnQTgtG6)
*   [The Squiggle Language Homepage](https://www.squiggle-language.com/)
*   [Squiggle Newsletter (Part of the QURI Newsletter)](https://quri.substack.com/t/squiggle)

Functionality
-------------

**Squiggle (the language)**

*   Write functions that accept and return probability distributions. Squiggle generates automatic plots for these.
*   You can provide [explicit ranges for functions](https://raw.githubusercontent.com/quantified-uncertainty/squiggle/main/packages/website/allContent.txt). This helps with the visualization, and ensures they won't be called outside that range. Like, appleStockPrice(t:\[2023, 2060\]).
*   [Custom plots](https://www.squiggle-language.com/docs/Api/Plot) for distributions and functions. Scales include linear, log, and symlog (like log, but with support for negative values). Symlog scales accept a parameter constant that you can use for adjusting the scale.
*   Make custom [tables](https://www.squiggle-language.com/docs/Api/Table

... (truncated, 12 KB total)

Resource ID: 15ed2e53e53d1be3 | Stable ID: sid_DBGZ0JdtyV