For the longest time, Mozilla was synonymous with the Firefox browser, but in recent years, Mozilla has begun to look beyond Firefox, especially as the importance of its browser continues to decline. In recent years, Mozilla has also started making startup investments, including in its Mastodon Mammoth client, for example, and it acquired Fakespot, a website and browser extension that helps users spot fake reviews. The organization also launched Mozilla.ai to bring more of the open source ethos to the AI space. Not surprisingly, AI is what the organization is focusing on right now. Indeed, when Mozilla released its annual report a few weeks ago, it also used the moment to add a number of new members to its board — the majority of whom are focused on artificial intelligence.
Late last month, I sat down with Mozilla president and CEO Mark Surman to discuss what’s next for Mozilla — and what that means for fans and Firefox.
“For the last year and a half, we’ve been focused on making a pretty dramatic change at Mozilla — to make it not just more than the browser but more of our kind of activist persona and building a kind of portfolio that regulates us — and regulates others—to go and take our values into the age of artificial intelligence or the next age of the Internet, however you want to talk about it.”
Mozilla AI
Mozilla released Mozilla.ai right around the time GPT-4 was released and the first Llama models became widely available. Surman described this as a “moment of focus” for the organization. “Mozilla AI, which had a broad mandate about finding open source, reliable AI opportunities and building a business around them. Quickly, Moez [Draief]who runs it, talked about how we can tap into the growing snowball of big language open source models and find a way to accelerate that snowball but also make sure it’s rolling in a direction that fits our goals and fits the zone of our wallet.”
While Mozilla made some headlines for launching its AI efforts, we haven’t actually seen much movement in this area from the organization since then. Shurman told me that the leadership team had been planning these efforts for nearly a year, but as public interest in artificial intelligence grew, they “kicked it out the door.” But then Draief almost immediately put it into stealth mode to focus on what to do next. “At the high level, where we position ourselves to make it easy to use any of the big open source language models in a reliable, privacy-sensitive and accessible way.” Right now, Surman argued, it remains difficult for most developers—and even more so for most consumers—to run their own models, even as more open-source models are seemingly released every day. “What Mozilla.ai is really focused on is almost building a wrapper that you can put around any big open source language model to optimize it, build data pipelines for it, do it with high performance.”
Exactly what that will look like remains to be seen, but it looks like we’ll be hearing a lot more in the coming months. Meanwhile, the open source and AI communities are still figuring out exactly what open source AI will look like. Surman believes that regardless of the details, however, the overall principles of transparency and the freedom to study the code, modify it and redistribute it will remain basic.
“Is it just the freedom to redistribute the final model? Is it the ability to study what’s inside? Is it to know what the weights are, to see what the data was? I think we are still working on all these questions. We’re probably leaning towards everything should be open source — at least in an intellectual sense. The licenses aren’t perfect, and we’ll be doing a bunch of work in the first half of next year with some of the other open source projects around clarifying some of those definitions and giving people some mental models.”
Surman believes that open source AI is a necessary ingredient to make the next era of the Internet open and accessible to all — but on its own, it’s not enough. With a small group of very well-funded players currently dominating the AI market, he believes the various open source groups should come together to collectively create alternatives. He likened it to the early days of open source—and especially the Linux movement—which aimed to create an alternative to Microsoft. Then, he noted, when the smartphone arrived, there were a few smaller projects aimed at creating alternatives, including Mozilla (and at its core, Android is apparently also open source, even as Google and others have built walled gardens around the actual user experience). However, these efforts were not so successful.
Surman seems optimistic about Mozilla’s place in this new age of AI, however, and its ability to use it to advance its mission and build a sustainable business model around it. “All that we are going to do is in the service of our mission. And some of it, I think, should just be purely a public good,” he said. “And you can pay for public goods in different ways, from our own resources, from charity, from people raising funds. […] It’s a kind of business model, but it’s not commercial, per se. And then the stuff we’re building around community AI hopefully has real enterprise value if we can help people exploit the models of big open source languages, efficiently and quickly, in a way that’s valuable to them and is cheaper from using open source EVERYTHING INCLUDED. That is our hope.”
What’s next for Firefox?
Where does all this leave the Firefox browser? Surman argued that the organization is being very cautious about bringing AI into the browser — but he also believes that AI will become a part of everything Mozilla does. “We want to implement artificial intelligence in a way that is reliable and benefits people,” he said. Fakespot is an example of this, but the overall vision is bigger. “I think what you’re going to see from us, over the course of the next year, is how do you use the browser to represent you and how do you build AI in the browser that’s basically by your side as you navigate the web?” He noted that an Edge-like chatbot in a sidebar could be one way to do that, but it seems he’s thinking more of an assistant to help summarize articles and perhaps alert you proactively. “I think you’re going to see the browser evolve. In our case, this is to further protect and assist you. I think it’s more that you use the predictive and compositional capabilities of these tools to make it easier and safer to get around the Internet.”
In the early days of Firefox, people moved away from other browsers because Firefox was much better at blocking annoying pop-up ads. Now, Surman argues, Mozilla needs to think about what pop-up blocking amounts to for today’s users. “The question we are now asking ourselves is: What is the pop-up blocker for the age of artificial intelligence? What are people really going to want that represents them and makes the internet experience better?”