The EU Commission has put forward a proposal for a harmonized regulation for the use of artificial intelligence. The proposal is both cheered and criticized. It will be ratified sooner rather than later, and the consequences are significant. The proposal itself is a thick pile of legal formulations, but here's a shorter and more approachable summary: The future of AI regulation in Europe and its global impact (cliffordchance.com)
The criticism is mainly directed towards two issues; 1) how the regulation will limit European competitiveness in the international AI race, 2) that there are still several essential grey zones and open questions to be tackled. The latter is what the proposal process is for; to get input and feedback on open issues. I trust the process and the efforts of some of the smartest heads in Europe to sort this out. The first issue, though, is worth reflecting upon; will a limitation on how AI may be used to influence or impact you and me as individuals limit the European competitiveness in the global AI race?
Before I reveal my current position on this question, I would like to invite you to join me on a quick mental journey; humans have since our very origin been sceptical towards those who are different (aka biased), we have made sure to take care of ourselves and those dear to us before sharing with others (aka egostic), and we have risen to the occasion when the opportunity to take control over resources have emerged (aka greedy). We, humans, have learned to behave in ways that make the tribe support us (aka influencing), and we punish those who go against the interests of the tribe (aka social control). All these features are related to the survival and progress of societies, we're all part of it, and mostly it has been for the best of our society. The trouble is, though, that history has shown us time and again that humans have a tendency to not really be all that nice to each other, consciously and unconsciously.
In a digital world where we introduce mathematical algorithms to replicate and automize human skills and behaviour, such as interpretation of images and sounds, understanding of languages, and the ability to reason, we are not only replicating the factual skills and good intentions, we are in fact also replicating the thoughtless and less attractive features of us humans. How did this happen? Simple. The data we train our algorithms on are a documentation of human behaviour, for good and for bad. The algorithms learn to identify patterns in the data, and use these patterns to predict future behaviour. The data become a mirror of our human oh-so-fallible behaviour. Then we blame the algorithms, or better, we blame the person who developed them, whilst completely forgetting how these algorithms learned to behave that way in the first place (hint: the mirror…). The examples are abundant; recruitment, criminal justice, social surveillance, credit scoring, etc.
We don't want to create machines that automize human thoughtlessness. We want to create machines that automize human brilliance. That's why we need the AI Act. Will it hamper competitiveness? Yes, if your drive is to create intelligent automization of human thoughtlessness.
However, there are so many challenges our there that need solving; green energy, environment, climate, utilization of the oceans, food production, mobility, production of various kinds, +++. Europe has amazing opportunities to take a lead on the digitalization of these industries, including the application of artificial intelligence.
So no, I do not believe the AI Act will hamper European competitiveness in the AI race. On the contrary. We may in fact help humanity to become a better version of itself.