Mozilla’s 2024 Rise25 Honoree Interview: Roberto Di Cosmo

My advice is to build a strong foundation in both science and humanities, think critically, debate ideas respectfully, and pursue impactful work that contributes to the greater good.

What is the biggest tech & society issue we are currently facing?

There are at least two on my radar: The growing fragility of our software infrastructure and the rapid advancement of artificial intelligence.

Open source launched an unprecedented era of innovation by providing freely available and modifiable building blocks for almost all technology needs, letting innovators focus on the new value added, instead of reinventing the wheel. But now applications rely on hundreds or thousands of external components, a software supply chain that has become increasingly complex and vulnerable, as acknowledged by recent regulation from the Executive Order on Cybersecurity to the Cyber Resilience Act. We must urgently secure the software infrastructure critical to our modern world.

Like software in the past decades, AI technologies are now coming into every aspect of our lives—from healthcare and education to finance and governance— and at an unprecedented pace. Generative AI models exhibit behavior indistinguishable from humans, easily passing the famous Turing test, raising concerns on by whom, how and for whom they are built. Models trained on vast historical datasets often remain proprietary due to costs and secrecy. This lack of transparency hinders reproducibility, bias identification, and quality assessment. This issue mirrors the challenges of the software supply chain, but with even greater stakes.

Addressing these intertwined challenges requires a concerted effort to build the tools and regulations for transparency, accountability, and security in both AI development and software infrastructure, ensuring that technological advancements benefit society as a whole.

How does your role help tackle thorny tech & society issues? 

As a professor, computer scientist, and champion for open-source and open science, my work centers on fostering transparency, accessibility, and collaboration in the digital ecosystem. Over the decades, I've championed open-source software, promoting, funding, and protecting vital initiatives for developers and society. For example, creating and leading the Mancoosi project almost twenty years ago to address the scientific challenge of solving complex package dependencies in software ecosystems, a crucial issue for the long-term stability of open-source infrastructure. 

Today my main focus is on Software Heritage, an initiative launched by Inria in partnership with UNESCO some ten years ago to collect, preserve and make available the collective digital knowledge encoded in software source code. Software Heritage is a crucial repository of source code, acting as a 'large telescope' into the intricate world of software development. It addresses key challenges in the software supply chain, including vulnerability detection, licensing compliance, and understanding software evolution. (An example: The Software Heritage Security project that uses the history and provenance of code to detect vulnerabilities and improve software resilience).

I've been a vocal advocate for transparency and accountability in AI, as outlined in the Statement on LLMs for Code, outlining principles for ethical, transparent AI development. The BigCode project followed these principles for the StarCoder2 model, demonstrating how to develop efficient models while maintaining data transparency and open-source code for research. This approach ensures that AI is a force for good for all and allows researchers worldwide to contribute and benefit from advancements.

How did your career grow, and what advice would you give to others wanting to be in a similar position?

My high school education in Italy was in classical studies, focusing on history, philosophy, Greek, and Latin, but I was already passionate about science. The money from a national award for translating Greek to Latin funded my first computer, a ZX80. I was drawn to computer science, but my humanistic background has always been key to my approach. I studied theoretical computer science and logical systems at the Scuola Normale Superiore in Pisa, which inspired me to pursue research and teaching.

After my PhD, I moved to France and joined École Normale Supérieure de Paris as an associate professor, later becoming a full professor at Université Paris Diderot (now Université Paris Cité). In 1998, my book on software monopolies helped spark the growth of the Free and Open Source movement in France. I advocated for open alternatives and saw the potential for innovation and democratization through FOSS and Open Access. This led to the founding of initiatives like the Free Software thematic group (now Open Source Hub), the Mancoosi project, and Irill, a center dedicated to the study of open-source software. In 2015, I took leave from teaching to focus on Software Heritage.

My advice is to build a strong foundation in both science and humanities, think critically, debate ideas respectfully, and pursue impactful work that contributes to the greater good.

What backgrounds or voices would you like to see more of in the Responsible Tech ecosystem?

It takes a unique blend of competencies is needed to tackle the complex challenges that we face when looking for “responsibility” in technology. A humanistic vision rooted in a humble but deep understanding of history and philosophy will help identify and navigate the multiple societal implications of new technological developments. A clear understanding of the power and limits of political action will help ensure that we set realistic and achievable goals for global regulations. A broad knowledge of the technology involved will help avoid unrealistic expectations and ground our strategies in what is technically feasible. As the Italian Enlightenment philosopher Giambattista Vico said, 'knowing is doing.' We should apply this wisdom by combining the theoretical knowledge of academia with the practical experience of startups. Academia, when freed from the “publish or perish” curse, plays a vital role by conducting in-depth, long-term research that helps understand the origin and perspectives of new technology. Small startups may bring fresh, innovative approaches, often working outside the constraints of corporate agendas. By supporting a balanced mix of technical expertise, ethical understanding, and practical political awareness, we can create hope to foster a responsible tech future that bridges competing value systems and serves the needs of all people.

What is your vision of a better tech future and how we can we move towards that vision?

My vision of a better tech future is grounded in the belief that technology itself is neutral—neither inherently good nor bad. Its impact is shaped by how we choose to use it. History has taught us that while each new wave of technological innovation can seem disruptive or even frightening at first, society eventually learns to adapt and integrate these innovations in ways that propel us forward. From the rise of computing to today's advances in AI, what was once viewed as a threat has often evolved into a tool that enhances human capability, rather than replacing it. I’ve seen this firsthand over several decades of technological progress. Each time a new tool emerged that was predicted to replace developers, it instead enabled them to build even more complex systems, ultimately increasing the demand for human expertise. I believe we’ll see a similar trajectory with AI and generative AI—while the pace will be faster and the impact more pervasive, humans will adapt and use these technologies to drive further innovation. The challenge lies in ensuring that the values and culture guiding the development and deployment of these powerful tools are rooted in humanism and responsibility.Even as technology advances, our values will ultimately determine how we use it. My hope is that humanistic values will prevail, allowing us to harness the potential of AI and other technologies to benefit society as a whole.

What does recognized as a Rise25 winner mean to you?

Being recognized as a Rise25 winner is a powerful affirmation of our mission at Software Heritage. For over 25 years, I’ve dedicated my personal and professional life to promoting, developing, and preserving Open Source and Open Science. This award reflects the importance of these values in today’s world, where software is not just the fuel of innovation but the fabric that binds together all aspects of our digital lives. As new technologies like AI emerge they bring new challenges. This recognition strengthens our resolve to address them with transparency and openness. We believe that the benefits of these innovations should be shared by all humanity, from the north to the global south, and not locked behind closed doors. Joining the Rise25 cohort is both an honor and a reminder that we’re part of a global community dedicated to making AI and software development serve society in a transparent, accountable way. This recognition motivates us to continue pushing forward, ensuring that the knowledge embedded in software remains accessible to future generations.