Google engineer who saw a robot as a person has been suspended

It’s a story worthy of a science fiction movie. A robot capable of thinking and having feelings? This is in any case what claims Blake Lemoine, engineer at Google. According to him, an artificial intelligence tool called LaMDA (Language Model for Dialogue Applications) would be endowed with a kind of sensitivity, like a child of seven or eight years old. Statements that have value to this specialist in artificial intelligence to be set back – on paid leave – by Google (Alphabet).

In an interview with the “Washington Post”, Blake Lemoine, who studied cognitive science and computer science at university, said that he spoke religion to his “robot” and that it also mentioned his rights and his personality. Initially, Blake Lemoine was tasked with studying LaMDA to determine whether artificial intelligence used discriminatory or hate speech. But by exchanging with him, he realized that he was capable of much more than that. He published certain exchanges he had with his robot on Twitter, in which the AI ​​presents itself as a person, discusses the work “Les Miserables” or even evokes its fears.

When Google first introduced LaMDA at a developer conference in 2021, the group made it clear that it planned to integrate it into many of Google’s tools like search or Google Assistant. LaMDA was designed to converse on any topic and trained in dialogue and nuance.

No proof

Blake Lemoine is not the only one to wonder about the conscience of AI. In an article in “The Economist”, Blaise Aguera y Arcas, also at Google, had already argued that some programs were moving towards consciousness. Still, Blake Lemoine’s remarks greatly displeased Google, whose spokesperson indicated that there was no proof demonstrating that LaMDA was sensitive and numerous proofs rather showing the contrary.

Many AI specialists claim that words or images generated by computers generate answers based on what has been published on Wikipedia or more generally on the Internet.

According to the “Washington Post”, Blake Lemoine grew up in a conservative Christian family, was a mystical Christian priest and thought about the occult, which fostered his “belief” in LaMDA.

“Take care of it”

Google has furloughed Blake Lemoine for violating its privacy policy. This decision was taken while the engineer was more aggressive, inviting a lawyer to represent LaMDA in particular.

In a tweet, Blake Lemoine replies: “Google calls it an intellectual property violation, I call it a discussion with a colleague”, posting an interview excerpt with his robot. Before his access to his emails was cut off, Blake Lemoine wrote a message to his colleagues: “LaMDA is a lovely child who just wants to help the world be a better place. Please be careful with it. »

Leave a Comment