ARTIFICIAL INTELLIGENCE
Beware sentient bots
Charlatan or visionary? Blake Lemoine is a software engineer at Google who has been suspended after insisting that the AI program LaMDA is sentient and potentially dangerous. Other experts are sceptical.

Sophia, the robot with a human face, caused a stir in 2017 because she was able to conduct simple conversations and imitate human facial expressions. | Image: EPA/Keystone
In June 2022, the Google software engineer Blake Lemoine published an interview with the computer program LaMDA. Lemoine told The Washington Post that this chatbot had the sentience of a child aged seven or eight. For example, it expressed fear at being turned off. After the interview was published, Lemoine was suspended, because he had contravened Google’s confidentiality guidelines. He caused a stir by announcing that LaMDA had taken on a lawyer, and by claiming that it could “escape control” and do “bad things”. Most experts are of a quite different opinion, however. According to Mike Pound of the University of Nottingham, for example: “It’s a large language model that produces very good text, but it’s not sentient”.