I look forward to moderating the Panel Session on AI Education and Skills at the High Level Conference on AI: From ambition to action on 15 September. For several reasons, but above all because of the urgency to address the educational component of this fast-paced socio-technical development. It is not only essential that we have the skills to build an AI socio-technical infrastructure. We also need the awareness and education to make good use of the opportunities that AI brings and to mitigate its potential risks.

AI systems are everywhere. Recommendation systems are informing our views on social media. Risk assessment systems are foreseeing and mitigating risks in manufacturing settings. Precision medical systems predict patterns in patient data. The list goes on.

At the same time, AI systems are ”nowhere” - because how many of us notice these systems in our everyday lives? How many know how they work? Understand their full potentials or their risks and ethical implications?

AI systems are transforming life as we know it, but we are not all empowered to develop, critically understand and use AI systems.

While the first digital divide is related to the access to digital services and systems, such as the more advanced AI systems, the second one emerges as a gap between those with the skills and education to build and/or make critical use of these tools and others who do not have such skills.

Just being connected, doesn’t mean empowerment per se.

As the UN's World Summit on the Information Society's Declaration of Principles' describes it: "Everyone should have the opportunity to acquire the necessary skills and knowledge to understand, participate actively and benefit fully from the information society and the knowledge economy. "

Now, we are only catching up with the skills, awareness, and education we need for the Digital Information Age. Even more advanced AI skills and awareness are the next frontiers we must conquer and it is evident that we have to move fast.

Digital, as well as AI, skills, and education gaps are symptoms of social divides. The pandemic has shown how social gaps and inequalities are reproduced and reinforced by divides in digital access and digital skills. Education gaps related to AI skills make such social divides more visible. For example, while technology allowed many of us to carry on with our work and education as usual during the pandemic, when connectivity, equipment, or skills are lacking it can become a barrier. While all the above has been and is, hopefully, only a moment of disruption of our everyday lives, we still see inequalities generally reproduced in cases such as women who are trained AI trained and employed in AI filled but still lagging behind.

Developing a European AI infrastructure requires due consideration of the changes needed in our education systems as well as in our lifelong learning skills.

It means, for example, that we include the technical, social, and cultural AI skills holistically across curricula in higher as well as primary and secondary education. It also means creating attractive research and working conditions to keep our AI talent in Europe. Europe has overall a good amount of AI specialists. Nevertheless, EU comparable companies have fewer AI specialists than respective US companies. Stanford University AI Index report from 2021, for example, states that the majority of the US AI Ph.D. graduates are from outside the US— but they choose to remain there. Europe needs the conditions for attracting and keeping talent and it needs to equip the present and future workforce with a new set of human-centric AI skills.

Importantly, we’ve learned from the first wave of digital education in Europe that the skills set we need goes way beyond the technical capability to access and even to design technology. We also need to consider the less invisible and often taken-for-granted cultural resources and skills – the ethics and values-based conceptual frameworks - that are crucial for technological development and change.

If we want human-centric Trustworthy AI we have to ensure that our values take form in the systems we build.

We have to ask for it when we procure the systems, we have to nurture these human-centric skills in AI research and development. There are many ways for doing this, for example, as proposed in the EU White paper on Artificial Intelligence by, in very concrete ways, including the AI High-Level Expert Group's on Assessment List for Trustworthy AI into an indicative “curriculum” for developers of AI.

I am hopeful. We are on the right way. The skills dimension is an integrated component of the EU’s AI regulatory package and all EU Member States that agreed to adopt national AI strategies have equally integrated the skills component into their programmes. Generally, what we need now is a holistic and multidisciplinary approach with various education stakeholders and with the inclusion of educational institutions at all levels with the capacities to also upskill an adult workforce.

The above-mentioned aspects and much more are going to be discussed in our Conference Session on 15 September. What are your thoughts on AI Skills and Education?

I will be happy to receive your comments in the thread below!

Find out more on the High-Level Conference on AI: From Ambition to Action