'Stopping AI because it carries risks? That would be like banning cars because they cause accidents'

For several months, calls to "slow down," "regulate" or even "boycott" artificial intelligence have multiplied. Pope Leo XIV has warned against its effects on human dignity, American students have booed Silicon Valley executives, unions are worried about jobs, artists about creativity, teachers about learning. Everywhere, the same fear that AI will ultimately rob us of our humanity is cropping up. And athough this concern is legitimate, it obscures the real question.

The public debate now features two distinct camps: those who support acceleration and those who want to hit the brakes, as if the issue came down to a question of speed. Yet the real issue is not speed – it is trust. And trust is not a matter of technology; it is a matter of institutions.

Let us start with the obvious: Those in favor of "slowing down" underestimate to what extent algorithms already outperform us. In healthcare, finance, logistics, scientific research and many other fields, they already surpass humans in detection, diagnosis and optimization. Psychologist Paul Meehl's work – developed even further by Nobel Prize-winning economist Daniel Kahneman – established long ago that humans are imperfect decision-makers, subject to biases and random errors. Algorithms are, in many cases, more efficient and more consistent. Rejecting AI means missing out on potential benefits.

It also means fighting the wrong battle. The main argument for "slowing down" is a kind of precautionary principle – namely, it would be dangerous to trust "black boxes" whose workings we do not understand. Yet every day, we trust technologies we do not understand. We take medicines without knowing their biochemistry, we board airplanes without understanding how they fly, we drive cars that can kill.

If we place our trust in these products or machines, it is not because they are simple, transparent or harmless. It is because, over decades, we have built a genuine infrastructure of trust around them: laws, regulations, oversight authorities, certification procedures, audits, safety requirements, incident reporting systems, clearly defined responsibilities, post-accident investigations, insurance, legal precedents.

You have 62.11% of this article left to read. The rest is for subscribers only.