Conclusions of AI Forum 2018

The ongoing fierce international competition requires a shared vision of Europe that is competitive in the age of artificial intelligence – while building on the strengths and values of our continent. To enhance the European discussion on artificial intelligence, the Ministry of Economic Affairs and Employment of Finland and the European Commission organised the AI Forum 2018 in Espoo, Finland, in October 2018. The main conclusion of the Forum are below.

Capitalizing the potential of AI for European industry

AI drives industries’ competitiveness into next level, an opportunity no European industry can miss. Coordinated EU strategy, functioning Digital Single Markets and vast data resources are all needed for boosting business in the age of AI. Competitive European companies are also global companies, and should have a strong advantage on an international scale.

Co-operation and learning in ecosystems between companies in similar sectors is important, while at the same time we should facilitate learning between AI tech companies and application companies. Systemwise, the goal should be interoperability rather than aiming for one European platform. Increasing interoperability would create demand and make Europe more innovative. Streamlining interoperability, however, would require seamless cooperation even between competitors.

Instead of selecting priority industries to be accelerated, the European Commission could define problems that we should fix together with business and research communities. These problems should not be tech specific, but rather impact-driven, such as creating less waste or safer transport. We should show the value of AI to citizens to create trust and a predictable environment, with focus on solving big, significant problems and creating well-being.

AI integrated into policy-making, leveraging cooperation

AI is a general-purpose technology that should be integrated into all policy-making, the approach should be horizontal in covering all sectors. To further this end, administrators with traditional backgrounds, such as legal or economics, would need to better understand the new environment, opportunities and challenges brought by AI.

At the same time, policy makers should not need to figure out the AI competiveness issues alone. In the networked world, the core is people collaborating. We could create a cross-sector group of European industrialists to facilitate the dialogue on European competitiveness in the age of AI. Europe is already set up like a tight network, and that is a quality we should leverage.

Competitive advantage through AI ethics

Ethics should not be seen as a stumbling block, but an enabler providing us a competitive advantage that shapes innovation and technology. Successful implementation of ethics requires combining ethics into systems and algorithms, while ethics in design makes the choices made in development and training phases visible.

Algorithmic transparency and explainability of algorithms should be encoded into both the design and the processes as many models in use in AI today, too complex for even developers to understand. Transparency transparency is seen important society-wide but alone will not generate trust. Reducing AI ethics into ethics washing should be avoid-ed. We should make active choices rather than let things happen, at the same time, deciding on the right time to regulate is challenging. It is worth remembering that there is already a lot of relevant regulation in place.

Artificial intelligence changes the learning life

In the AI era, skills will be more essential to European industry as ever before. We need highly skilled developers. In addition to learning about algorithms, useful digital skills include communication, social skills and humanities.

When earlier the focus has been on degrees, we now need a greater variety of learning forms and increased modularity, such as short courses and online resources. Most importantly, people should learn to learn. More resources and new types of public-private partnerships are needed to educate people who are in the working life in an anticipatory way, instead of based on immediate needs.

Scaling research excellence and attracting rising stars

Scaling of research excellence calls for building top tier research centers, creating sand box environments, providing dynamic funding models and encouraging risk-taking at the highest level. World-class research centers cannot be everywhere, and it is important to make targeted and specialized research.

To attract the best research talent to Europe, we need to understand how to motivate researchers. Besides compensation, we should give them the most interesting research challenges and coolest tools. Europe is also lacking big corporations that would fund research. Further, seeing failure as an essential part of learning should be a crucial part of a successful research funding approach.