Distill is a new, web-based, peer-reviewed journal for academic research on machine learning. As YC Research put it in their announcement:

The web has been around for almost 30 years. But you wouldn’t know it if you looked at most academic journals. They’re stuck in the early 1900s. PDFs are not an exciting form.

Distill is taking the web seriously. A Distill article (at least in its ideal, aspirational form) isn’t just a paper. It’s an interactive medium that lets users – “readers” is no longer sufficient – work directly with machine learning models.

Ideally, such articles will integrate explanation, code, data, and interactive visualizations into a single environment. In such an environment, users can explore in ways impossible with traditional static media. They can change models, try out different hypotheses, and immediately see what happens. That will let them rapidly build their understanding in ways impossible in traditional static media.

I have to say I’m pretty excited about this and really hope it catches on in other areas as well. They provide many tools and workflows for researchers to publish their work in a web format. The peer review uses GitHub for an open process: you create a new repository, develop your paper in it and submit it for review to Distill.

This is also my only gripe about Distill: You can keep your repository private while it’s still under review (for publishing it needs to be public and CC-BY-licensed), but due to their reliance on GitHub this will of course require a paid account there. While 7$ / month won’t be too much for most (but maybe not all) researchers, it’s still a bit disappointing for an otherwise so open project.