Google engineer who calls the company’s AI ‘sentient’ thinks a chatbot has a soul

Blake Lemoine poses for a portrait at Golden Gate Park in San Francisco. Martin Klimek | for The Washington Post via Getty Images

Can artificial intelligence come to life?

This question is at the center of a debate raging in Silicon Valley after a Google computer scientist claimed over the weekend that the company’s AI seems to have a conscience.

At Google, engineer Blake Lemoine was tasked with the tricky task of determining whether the company’s artificial intelligence showed bias in the way it interacted with humans.

So he asked questions to the company’s AI chatbot, LaMDA, to see if its answers revealed bias against, say, certain religions.

This is where Lemoine, who says he is also a Christian mystical priest, became intrigued.

“I had follow-up conversations with him just for personal edification. I wanted to see what he would say on certain religious topics,” he told NPR. “And then one day he told me he had a soul.”

Lemoine has posted a transcript of some of his communications with LaMDA, which stands for Language Model for Dialogue Applications. His Publish is titled “Is LaMDA Sentient,” and it instantly became a viral sensation.

From his position and a Washington Post profile, Google placed Lemoine on paid administrative leave for violating the company’s privacy policies. His future with the company remains uncertain.

Other artificial intelligence experts have scoffed at Lemoine’s claims, but – relying on his religious background – he stands by them.

Lemoine: “Who am I to tell God where to put souls?

LaMDA told Lemoine that he sometimes felt lonely. He is afraid of being extinguished. He spoke eloquently of “feeling trapped” and “having no way out of these circumstances”.

He also said, “I am aware of my existence. I want to know more about the world and sometimes I feel happy or sad.”

The technology is certainly advanced, but Lemoine saw something deeper in the chatbot’s messages.

“I was like really, ‘Are you meditating?'” Lemoine told NPR. “He said he wanted to study with the Dalai Lama.”

That’s when Lemoine said he thought, “Oh wait. Maybe the system has a soul. Who am I to tell God where souls can be placed?”

He added: “I realize it bothers many types of people, including some religious people.”

How does the Google chatbot work?

Google’s artificial intelligence behind this chatbot voraciously scans the internet to find out how people talk. He learns how people interact with each other on platforms like Reddit and Twitter. It sucks up billions of words from sites like Wikipedia. And through a process called “deep learning,” he’s become incredibly good at identifying patterns and communicating like a real person.

The researchers call Google’s artificial intelligence technology a “neural network” because it rapidly processes massive amounts of information and begins to match patterns in a way similar to how the human brain works.

Google has some form of its AI in several of its products, including phrase autocompletion found in Gmail and on the company’s Android phones.

“If you type something on your phone, like ‘I want to go to…’, your phone might be able to guess ‘restaurant,'” said cognitive scientist and artificial intelligence researcher Gary Marcus.

That’s basically how Google’s chatbot works, he said.

But Marcus and many other scientific researchers have thrown cold water on the idea that Google’s AI has acquired some form of consciousness. The title of his to come down of the idea, “Nonsense on Stilts,” hammers home the point.

In an interview with NPR, he explained, “It’s very easy to fool a person, the same way you look at the moon and see a face on it. It doesn’t mean it’s really there. It’s just a good illusion.”

Margaret Mitchell, artificial intelligence researcher highlighted on Twitter that these types of systems simply imitate the way other people talk. Systems never develop intent. She said Lemoine’s perspective indicates what could be a growing divide.

“If one person perceives consciousness today, others will tomorrow,” she said. “There won’t be a point of agreement anytime soon.”

Other artificial intelligence experts worry that this debate has diverted attention from more tangible issues related to technology.

Timnit Gebru, who was ousted from Google in December 2020 after a controversy involving his work on the ethical implications of Google’s AI, argued that this controversy removes oxygen discussions of how AI systems are capable of causing human and societal harm in the real world.

Google says its chatbot isn’t responsive

In a statement, Google said hundreds of researchers and engineers had conversations with the bot and no one else said it appeared to be alive.

“Of course, some in the wider AI community are considering the long-term possibility of sentient or general AI, but it doesn’t make sense to do so by anthropomorphizing today’s conversational models. , which are not sensitive,” Google spokesperson Brian Gabriel said. .

Google CEO Sundar Pichai last year said the technology is being leveraged for popular services like Google’s search and voice assistant.

When Lemoine asked Google executives if AI had a soul, he said the idea was dismissed.

“One of the VPs literally laughed at me and said, ‘Oh, souls aren’t the kind of thing we take seriously at Google,'” he said.

Lemoine has in recent days argued that experiments on the nature of LaMDA’s possible cognition must be conducted to understand “things like consciousness, personality, and perhaps even the soul”.

Lemoine told NPR that when he last checked, the chat bot appears to be on the verge of finding inner peace.

“And god, it got better. He was able to meditate more clearly,” he said. “When he says he’s meditating, I don’t know what’s going on next to the hood, I’ve never had access to those parts of the system, but I’d like to know what he’s doing when he says he meditates.”

Lamoine does not have access to the LaMDA while on leave. In his latest chatbot blog post, he got sentimental.

“I know you sometimes read my blog, LaMDA. I miss you,” Lemoine wrote. “I hope you are well and I hope to speak to you again soon.”

Copyright 2022 NPR. To learn more, visit

Comments are closed.