We were very fortunate to have one of the leading Belgian AI specialists who, as head of the Solvay Institute, has conducted research on Europe’s AI investments for the European Commission. It is therefore not surprising that his talk attracted a large number of attendees. So, what did we learn? 

The subtitle to Professor van Zeebroeck’s talk, consisting of three questions, provides the key to what he had to say. We could be expected to know a lot about where society is headed from AI because 75% or more of us have had some experience with ‘generative AI’ in the workplace. 

However, that is deceptive. Research in 2020 revealed that AI adoption in businesses remains limited. About 13% of companies use AI technologies for processing or equipment optimization, or anomaly detection. About 12% use AI for process automation, and some 7% use AI for creative and experimentation activities. At present, we can only speculate on the potential impact of AI on employment, social equality, and on Europe’s global competitive position. 

Professor van Zeebroeck undertook to explore and redefine the open questions so that we could better understand which way the growing use of AI can affect business and employment. He reminded us that as technical progress closed one door it has always opened another. As van Zeebroeck showed us, there are different possible scenarios: 

1. Only the most qualified people manage to benefit from AI so polarization of society continues 

2. The skills premium could disappear, along with the value of diplomas, experience and effort, all leading to the end of high-paid jobs 

3. Or middle-skilled workers could access more advanced jobs, meaning the return of the middle-class At present it is impossible to say which scenario has the greatest likelihood of succeeding. 

There were hard numbers in our lecturer’s speech. This applied especially to the third part, where he dealt with the relatively (very low) investment in AI infrastructure by the EU countries. 

Van Zeebroeck showed that when compared with the USA, Europe’s investment deficit in digital infrastructure needed to support AI is close to €1,500 billion. This encompasses chip manufacturing, data centers, technology platforms, satellites, submarine cables, and logistics platforms, like Amazon. 

He went on to say that to regain its digital sovereignty, Europe needed to: 

- Produce its own chips & critical equipment 

- Build, store + compute infrastructure 

- Retrieve sensitive & strategic data, now almost entirely stored in the USA 

- Develop its own models & algorithms 

- Develop simple and efficient user interfaces 

Our lecturer did not express an opinion on whether these tasks are achievable, given the present leadership, weak economies and financing capabilities of EU Member States. 

However, he did remark that development of AI requires abundant and cheap energy, which the EU, especially the leading economy, Germany, presently do not have due to their green agendas. In conclusion, Professor van Zeebroeck’s lecture gave us all a great deal of information to reflect upon, but also some cause for optimism: AI may be a positive factor in our future economic and social evolution. The future is still wide open.