Experimenting with local alt text generation in Firefox Nightly – Mozilla Hacks - the Web developer blog

When I started working on AI at Mozilla two years ago, I was a Python developer with a background in web services and three months of machine learning experience from working on the Nuclia DB project. I was not someone who had trained models from scratch or built production ML infrastructure. Today, Firefox ships multiple AI features that run entirely on-device, and I helped build the infrastructure that makes that possible. This is a retrospective on what we accomplished and what I learned along the way.

Image generated by DALL-E in response to a request for a photorealistic image of a fox standing in a grassy landscape. Firefox 130 introduces automatic alt

I gave an introduction to the Firefox AI runtime and WebExtensions at a French local conference this month. This article is a loose transcript of what I said.

TL;DR: As Mozilla moves to make Firefox an AI browser, people are looking at other options. Some people are rediscovering Waterfox, a browser that has been around for a decade from independent developer BrowserWorks. In this post, I interview the founder of Waterfox - Alex Kontos, and we discuss Waterfox’s history and look towards its future. We also talk about how Waterfox thinks about AI in the browser.

Context: Improving Alt Text for Firefox

Image generated by DALL*E We’re shipping a new API in Firefox Nightly that will let you use our Firefox AI runtime to run offline machine learning tasks

What if you could call a method, from the JavaScript code of a webpage, and get AI-generated text from a language model that's running in your device? That's what Google has started to experiment with. Read on for more details.