The program document determined the narrative of the conference:
„As humanity becomes increasingly data dependable, it is necessary to make sure that the data it stands on reflect the values inherent to a fair, ethical, and economically sound society. … Data in itself is devoid of any value; it’s the way in which data is selected and used that can dictate how machine learning and other systems that feed on data, operate in economic, moral, or ethical terms. The question then arises: how can the systems and devices that use data in the data economy obtain economically, morally and ethically robust data?”
The conference started with the presentation of prof. Andres Guadamuz from Sussex University, who challenged the participants to differentiate artificial intelligence and man-made art and music. The audience had dificulties to see the differences between the two and it was a practical and efficient way to showcase the challenge of future IPR regulation.
The trust session highlighted the risks and benefits of using data for commerce and health sectors, not only for primary beneficiaries but also for the ecosystem around them. The morality and ethics panel had a wide scope, including the impact of data economy on the human rights in the future. A keynote was given by professor Anselm Kamperman Sanders on the „The intersection of intellectual property and data in the United Nations’ World Economic and Social Survey 2018”. It was a very inspiring account on how to deal with Risks and Benefits.
Economic value of data has a huge potential value for all sectors. This an opportunity we should be taking advantage of, but that should be regulated. Especially for data intensive sectors, regulation is a prerequisite to reduce uncertainties. It was a pleasure to be in the final panel with four talented students (Gaia Lisi; Eisa Rahimi; Virginia Debernardi; Bert Brookfield-Hird), with Prof Guadamuz and prof Ana Ramalho to discuss the issues from different perspectives. We agreed that Data is a new sector where regulation must happen fast, to enable European enterprises and citizens to harness the benefits.
However, legislation alone will not solve it. Due to human biases, datasets and data-analysis carried out by humans will resemble these biases. This is always an issue, but AI technologies amplify any bias and if they are not corrected in due time, it will have severe implications on the technology. Thus, certain issues have to be addressed at the expert level. We also discussed the global race. As a law-maker, I highlighted that while the EU might be lagging behind China, and the US on R&D spending, the EU is a strong exporter of legislation. The EU has been at the forefront of addressing legal issues, and has not been afraid to set standards high for companies. For example, the GDPR is an exemplar for other countries and regions in the world, while companies also praise it, notwithstanding the complexity of adopting it.
Bert concluded “The possible uses for data are constantly expanding and seemingly endless. However, great care must be taken to avoid the potentially huge benefits that data can bring to our society from being outweighed by the consequences. We must try to avoid this. Interdisciplinary events such as the value-less data symposium are key to ensure this and also so that we understand the issues before we legislate on them.”