
Camembert or Pie Charts – Why AI Translation Fails
Camembert or pie charts? AI recognizes patterns but misses meaning. What seems funny on Duolingo becomes troubling when machines judge where humans should understand.

Camembert or pie charts? AI recognizes patterns but misses meaning. What seems funny on Duolingo becomes troubling when machines judge where humans should understand.

When it comes to business ethics, AI companies ignore the most basic concepts linked to accountability, supply chain responsibility and product safety. Yes, AI companies create groundbreaking innovation. But that comes with the responsibility to ensure that what they do serves humanity, not the other way around.

In the context of art and creativity, AI touches on two fundamental dimensions of ethics: the question of the meaningfulness of life and the question of justice. Talking about democratisation in this context produces one thing above all: a lot of hot air.

AI in publishing was the topic of a debate, organized by Wiley. How should publishers respond to advances of Big Tech? How can AI make research more accessible? And do we really need to reinvent the wheel when talking about accountability in the age of AI?

AI is predominantly developed by men. It is also used more extensively by men. And in key datasets, men are overrepresented. This won’t do: it’s time for women to step in! That was my call to action at the Business Day for Women in Vaduz.

We praise AI for detecting patterns we can’t see. But how free are we if we stop thinking for ourselves? My TEDx talk explores what we risk when we delegate our judgment to machines.

UNESCO Forum invited me as a speaker to share my thoughts on the Covid-19 crisis. The pandemic has sparked fundamental ethical debates. Think of the terrifying reports from hospitals in Italy in Spring 2020. Intensive care units were overrun with patients. There were not enough ventilators. And suddenly we asked ourselves: What is the value…