Ethical and open artificial intelligence. Mozilla’s proposal (and bet).

Listen to the audio version of the article

The Mozilla Foundation is the good side of the web. Forever. We owe them the birth of the Firefox browser that today maybe few know but taught us the value of data transparency, privacy and open software. In fact, Mozilla is and truly remains the last (red) panda on the web. The only independent entity that has been thinking, producing and distributing technology for 14 years. The example of an alternative way of understanding the web to big names like Facebook, Google and Microsoft that move with business logic. In an interview with Il Sole 24 Ore, the president of the organization Mitchell Baker said: «The market, competition and commerce linked to search engines are the true driver of innovation. But the network is not just numbers and business models. There is a layer, a dimension of the internet which maintains a social and civil value. That’s what we wanted when we created Firefox, to create a place where the most important thing is you. For us it is the sovereignty of the user».

These words from about ten years ago are particularly topical today with the advent of generative artificial intelligence which presents itself as the most important system innovation since the days of the app store. That’s why we like the announcement of a new project from Mozilla to create ethical and open artificial intelligence regardless. Even if it won’t do anything. Or it will end up like Firefox did today. For now, very little is known about what they intend to do. We know that Mozilla will invest $30 million to launch Mozilla.ai, a startup with a mission to build open source, reliable AI in line with its founding principles. The goal is to build AI apps and products within an independent, decentralized ecosystem.

The currently planned team has 25 people, including engineers, scientists and product managers, to work on large language models. not only to develop products but also to understand how to make the processes that govern the algorithms transparent. Mozilla’s intent doesn’t appear to be demonizing current AI projects. But the timing is interesting.

It comes just as the two biggest names in AI for the general public, Microsoft and Google, are pushing their respective initiatives. The first presented the new Office suite enhanced by artificial intelligence, under the name of Microsoft 365 Copilot, while Big G made the Bard chatbot available in a preview version to an audience of selected users in the United States and the United Kingdom United Kingdom, with the aim of extending the trial to other countries in the coming months. Bard and Bing are obviously owned and operated by two of the biggest digital giants. The intention is evidently that of producing “closed” products for the market. In the beginning, OpenAI was also born with a logic not far from the Mozilla Foundation. Just like a non-profit artificial intelligence research organization with the aim of promoting and developing artificial intelligence. It promised to be the kind of open ecosystem Mozilla was talking about. But then he changed his mind and started sharing much less information about his models and training data. Also for this it is good that there is Mozilla.