Artificial intelligence (AI) has a major influence on how people deal with digital systems and services. What should society’s attitude be to this digital transformation? How can we find the right balance between freedom of action and the functional added value of AI? How can we safeguard public values, our fundamental rights and democratic freedoms? And how can we make sure that everyone benefits from the positive effects of AI?
Using AI raises a lot of questions. AI is a complex system technology that is capable of adapting to new situations. As a result, it is not always possible to predict what the effects of the deployment of an AI system will be. That means that many questions can only be answered as we go.
It is not enough to focus on human values and ethics in the development phase only, and then test whether a system really did what it should have done afterwards. There must be constant realignment between what humans want and what an AI system does and can do. There is a constant tension, requiring a careful approach to developing and maintaining human-centric AI.
Together with all the other parties involved, the Netherlands AI Coalition (NL AIC) is constantly looking for ways to learn and discover the best and most desirable AI solutions. There are already some partnerships where great things are happening in this area, such as the current development of the ELSA Labs, based on the ELSA concept, which is giving human-centric AI an extra boost. ELSA stands for Ethical, Legal and Societal Aspects. Not a new concept in itself, although the same cannot be said of the specific approach devised by the NL AIC. The ELSA Labs can be seen as social co-creation environments where we – together with the public, the commercial sector, centres of expertise and governmental authorities – investigate how AI can help shape our society.
The Human-Centric AI working group does more. We are trying to increase AI awareness among Dutch people. We are mapping out the new understandings and trying to bring the right parties in contact with one another. It is important to learn from each other so that the development of human-centric AI can be accelerated. With this objective in mind, we also have set up a platform for ethics that parties can come to when they are going to develop a new AI solution and want to make sure that the system or application does what they want (and will keep doing so).Additionally, the working group actively supports another initiative, the LegalAir platform, where people can find specific information, download template documents and ask experts specific questions about legal issues.
To get the kind of human-centric AI that we want in the Netherlands, it is important that the public have an influence on it. This means that they need to have enough information to be able to make the right choices when using AI in their daily lives. They should also be involved actively in developing new AI services so that they can say at an early stage what they think is desirable and what is not. With this goal in mind, the NL AIC is going to start an interactive dialogue with a million members of the public through the AI Parade. An approach as a co-creation of the Human-Centric AI and Culture and Media working groups, with the National Library of the Netherlands as an important partner. Twelve libraries from all over the Netherlands are taking part.
In the Human-Centric AI working group, we are closely involved in the initiatives mentioned above. We collect together the various insights, acquired for instance in all the ELSA Labs, think of possible solutions and validate them. We formulate widely accepted norms for the ethical, legal and social sides of AI systems and services. We then test those norms against international AI frameworks, preparing for future European regulations on AI as we do so. We applaud the fact that such regulations are being worked on and provide feedback to Brussels, yet at the same time we know that regulations alone are not enough to achieve the desired human-centric AI systems and applications. That will require the active and constant effort of multiple parties, in particular clients, developers, AI experts and users – not just in the Netherlands, but also internationally. We are working closely with other NL AIC working groups and there is clear interest from abroad in the way we are doing things in the Netherlands.
If you would like to know more about the ELSA concept, the various ELSA Labs, the dialogue with the public or the approach that the Human-Centric AI working group has adopted for ethical and legal problems, please contact Náhani Oosterwijk or one of the contacts listed. You can also take a look at the position paper on ‘ELSA Labs for Human-Centric Innovation in AI’ or the Human-Centric AI Manifesto (in Dutch).
If you’re interested in this topic and would like to be actively involved, you can become a member of the NL AIC and benefit from our Human-Centric AI expertise and network, and other relevant AI themes.