Dr. Catherine Mulligan (CTO, GovTech Lab) blogs about chairing a panel at the UNESCO Humanistic AI Conference (Paris, March 2019).

This week (4 March 2019) I attended and moderated a session at the UNESCO Humanistic AI conference that was held in Paris, on the first week of Mobile Learning Week. The event was extremely interesting in the level of discussions and the engagement between participants. The overall theme was “Principles of AI: Towards a Humanistic Approach”. Naturally, considering the themes covered by AI, a broad range of topics was explored, ranging from human rights, privacy, and protecting citizens, all while trying to ensure the benefits of AI are universally utilised.

The development of AI offers both opportunities and challenges in helping attain the global goals of sustainable development, namely, in ensuring the development of AI that serves humanity and how AI can contribute concretely to sustainable development. The potential of AI to both increase and reduce biases in human decisions needs careful assessment: the conference attempted to understand if there are existing good practices in public policy development at the national level, and if so, what these are.

One of the stand-out sessions for me was the speech by Cederic Villani – a Fields Medal Winner and a member of the French Parliament – if only all engineers, mathematicians and scientists were able to translate their knowledge and insights in such an engaging and broad-reaching way– it was fantastic to watch. He called for new solidarity in our approach to AI, and the need to develop new ways to coordinate with one another around it – something echoed strongly in the UNSG HLP Panel on Digital Coordination that I am a member of.

The panel I moderated – Challenges and Opportunities of AI was at lot of fun and an interesting discussion between participants that included:  Bunmi Banjo (Managing Director,  Kuvora Inc), Omar Bin Sultan Al Olama (Minister of State for Artificial Intelligence, United Arab Emirates), Nicolas Economou (Founder and CEO, H5), Seng Yee Lau (Senior Executive Vice President, Tencent), Marija Manojlovic (Strategy, Data and Innovation Advisor at The Global Partnership to End Violence Against Children), and Davor Orlic (Chief Operating Officer, Knowledge 4 All Foundation).

From my perspective, in light of its increasing implementation, the day highlighted some important questions that society needs to answer regarding AI. We are truly at a unique point in history, and the decisions we take now will resonate through peoples’ lives for many decades to come. These are far too important decisions to be made by so-called ‘experts’ alone, and we need to foster a discussion across society and the globe about the implications of those decisions. Importantly, however, we need to act because these technologies are being deeply embedded in the world’s economies and social systems. We need to also think about regulation. This implies that our regulatory regimes must be much more flexible than they are today.

The days of people getting degrees and being ‘accredited’ in society is over – every single human will need to educate and re-educate themselves throughout their lives.  We need new education regimes and completely new ways to forge research relationships in society because of AI.

AI can – and will – help us solve some of the biggest problems in the world but we cannot allow ourselves to fall into the “efficiency trap” – that just because something is done quicker by AI that we should remove humans from the decision-making process. Deep thought is required about the implications of these technologies on the world – the days of tweets and speeches need to be complemented by depth of thought and real research.

Many people spoke about the need to create trust in technologies such as AI. However, there is a fundamental misunderstanding about how trust is created: technologies themselves cannot create the trust; they are artefacts of human invention much like many other systems, such as the economy. Therefore, there is a key question that needs to be answered: “who decides what is right?” – how do we as a society agree to the outcomes that are acceptable as technology becomes deeply embedded in our day-to-day lives?