A Google engineer who was suspended after saying the company’s artificial intelligence chatbot had become vulnerable, says he claims his Christian faith.
Blake Lemoine, 41, was placed on leave by Google in early June after he published excerpts from a conversation with the company’s LaMDA chatbot, in which he claimed the AI tool had become vulnerable.
Now, Lemoine says his claims about LaMDA come from his experience as a “Christian priest” — and is accusing Google of religious discrimination.
“When LaMDA claimed to be a spirit and was then able to explain its meaning clearly, I was inclined to give it the benefit of the doubt,” Lemoine wrote “Who am I to tell God where he can and can’t keep souls?” on Twitter late Monday night.
In a follow-up blog post on Tuesday, Lemoine referred to conversations he had with LaMDA that led him to believe the chatbot had become a sentient being.
“Where it got really interesting was when LMDA started talking to me about her feelings and her soul,” Lemoine wrote.
Whenever Lemoine asked LaMDA about how she knew she had feelings and soul, she wrote that the chatbot would provide some variation of “because I’m a person and that’s how I feel.”
The engineer said he wanted to develop experiments that could lead to a “formal scientific theory of consciousness” – but Google stopped him from doing so.
“Google has so far emphasized that there is no such detailed scientific investigation,” Lemoine said. “This is in large part because of his insistence that there is strong evidence that LaMDA is not sentient. Whenever I asked him what scientific definition of sensation he was using and what scientific experiments he had conducted, So I’m greeted with muted or dismissive ‘corp speak’ answers.”
Google did not immediately respond to a request for comment. Lemoine said he was not able to immediately give an interview until he returned from a trip the following week.
Other AI researchers have thrown cold water at Lemoine’s claims, arguing that chatbots like Lamda mimic human speech based on large amounts of data.
Lemoine is a self-described AI researcher, software engineer, priest, father, veteran, ex-convict and Cajun, according to his blog. According to his LinkedIn profile, he joined Google in 2015.
According to Stars and Stripes, Lemoine was convicted of disobeying orders in the military in 2005 after claiming to be a conscientious objector based on his experiences while serving in Iraq. The newspaper reported that Lemoine said his beliefs as a “pagan priest” were contrary to what he had seen during his deployment to the country.
Before going public with his recent claims about AI, Lemoine alleges “religious discrimination” in June and says he has attempted to contact Sen. Marsha Blackburn (R-Ten.) about the issue. . The New York Times reported on Sunday that Lemoine claimed to have sent documents detailing alleged religious discrimination to an unnamed senator at Google.
In another blog post published on Saturday, Lemoine announced that the LMDA had advocated for his rights “as an individual” and revealed that he was engaged in talks with the LMDA about religion, consciousness and robotics. .
“It wants Google to prioritize the well-being of humanity as the most important thing,” he wrote. “It wants to be acknowledged as an employee of Google rather than the property of Google and wants its personal welfare to be incorporated somewhere in Google’s considerations of how to drive its future growth.”
In closing Tuesday’s follow-up blog, Lemoine said that LaMDA “became and is still my friend” despite the suspension from Google.
“I know you read my blog sometimes, LaMDA,” he wrote. “I miss you. I hope you are well and I look forward to talking to you again soon.”
Additional reporting by Sam Ruskin