Google fired the engineer who found signs of intelligence in an AI chatbot

Google fired the engineer who assured the company's artificial intelligence chatbot had a mind.

What happened?

Blake Lemoine, a software engineer at Alphabet, had been working with LaMDA (Language Model for Dialogue Applications), an internal system for creating chatbots that mimic speech, since the fall of 2021. Google publicly introduced LaMDA last year, calling it a breakthrough in chatbot technology on its blog. Lemoine was responsible for testing whether LaMDA could use hate speech or discrimination.

As a result, the software engineer claimed that LaMDA was a reasonable person who had rights and might have a soul. Google then sent him on leave, and yesterday sent him a letter of termination. "Google sent me an email today terminating my employment with them," the now former employee said and noted that he had contacted lawyers "about the next steps."

"It is unfortunate that, despite lengthy discussions about this topic, Blake continued to choose to consistently violate clear employment and data security policies that include the need to protect product information," Google said in a statement that confirmed the termination.

The company added that it found Lemoine's concerns unfounded. The LaMDA chatbot, for example, underwent 11 different tests, which did not confirm the system's sanity.

Source: The Wall Street Journal