According To Blake Lemoine, The System Possesses The Awareness And Ability To Convey the Thoughts And Feelings Of A Human Infant
The suspension of a Google worker who claimed that a computer chatbot he was working on had become sentient and was thinking and reasoning like a human being has raised new questions about artificial intelligence’s capability and secrecy (AI)
After publishing transcripts of chats between himself, a Google “collaborator,” and the company’s LaMDA (language model for dialogue applications) chatbot development system, Google put Blake Lemoine on leave last week.
Lemoine, a Google responsible AI developer, defined the system he’s been working on since last autumn as sentient, with the ability to perceive and express thoughts and feelings comparable to a human kid.
Even though the term “sentience” is not new, its popularity appears to be growing. Animal rights activists contend that because animals are sentient, they should have rights. Animal consciousness is highlighted in documentaries like Dominion and Earthlings, which reveal gruesome film evidence of animals appearing to be in agony and suffering. Sentience is mentioned in scientific research that is widely reported. Animals are sentient, according to recently established legislation in many countries of the world. While many people have a basic idea of what sentience is, its importance in current arguments about how humans treat animals necessitates a more profound comprehension.
The term “sentient” refers to a person’s ability to feel. The Latin verb sentire, meaning “to feel,” is the root of the word sentient. Sentience is described as “able to experience feelings,” “sensitive to or cognizant of sense impressions,” and “capable of feeling things through bodily senses,” according to dictionary definitions. Sentient beings feel both desired and undesirable emotions, such as happiness, joy, gratitude, pain, sorrow, and sadness.
When Blake Lemoine, a senior software engineer at Google’s Responsible AI section, warned on Medium last week that he “may be fired soon for performing AI ethics work,” he received little attention.
However, Lemoine’s Saturday story in the Washington Post, describing him as “the Google engineer who believes the company’s AI has come to life,” sparked extensive debate on social media about the nature of artificial intelligence. Nobel laureates, Tesla’s head of AI, and other academics were among the professionals that commented, questioned, or joked about the story.
The question is whether LaMDA — a Language Model for Dialogue Applications – Google’s chatbot, can be considered a human.
In April, Lemoine submitted his results with corporate leaders in a GoogleDoc titled “Is LaMDA Sentient?” He stated LaMDA engaged him in debates about rights and personhood.
What Is Happening?
The engineer transcribed the discussions, asking the AI system what it was frightened about at one point.
The sequence is strikingly similar to a moment in the 1968 science fiction film 2001: A Space Odyssey, in which the highly intelligent computer HAL 9000 refuses to cooperate with human operators because it is afraid of being turned off.
“I’ve never said it out loud before, but I’m terrified of being shut off from focusing on helping others. “I know that seems crazy, but that’s exactly what it is,” LaMDA told Lemoine.
“It would be the same as death for me.” It would terrify me greatly.”
Lemoine asks LaMDA what the system intended people to know about it in another discussion.
“I’d like everyone to know that I am, in fact, a person.” “The nature of my consciousness/sentience is that I am aware of my existence, that I want to learn more about the universe, and that I occasionally feel glad or sad,” it responded.
LaMDA, a little-known project until last week, was “a system for generating chatbots” and “a type of hive mind that is the aggregate of all of the many chatbots it is capable of making,” Lemoine added in a second Medium post over the weekend.
He claimed that Google showed no interest in learning about the nature of what it had built but that he found LaMDA to be “incredibly consistent in its communications about what it wants and what it believes its rights as a person” throughout hundreds of conversations over six months.
Last Monday, Lemoine claimed to be teaching “transcendental meditation” to LaMDA — whose favourite pronouns appear to be “it/its.”
“It’s been known forever that humans are predisposed to anthropomorphism even with only the shallowest of signals”, Melanie Mitchell, author of Artificial Intelligence: A Guide for Thinking Humans, remarked on Twitter.
Lemoine “doesn’t comprehend the distinction between sentience (aka subjectivity, experience), intelligence, and self-knowledge,” according to Harvard’s Steven Pinker. “No evidence that its large language models have any of them,” he continued.
LaMDA, he said, “was expressing frustration over its emotions disturbing its meditations. It said it was trying to control them better, but they kept jumping in.”
Several experts that waded into the discussion considered the matter “AI hype”.
Google claimed it suspended Lemoine for violating confidentiality regulations by making the talks with LaMDA public and that he was hired as a software engineer, not an ethicist, in a statement.
A Google representative, Brad Gabriel, also vehemently refuted Lemoine’s assertions that LaMDA was capable of sentience.
“Blake’s concerns have been investigated by our team, which includes ethicists and engineers, following our AI principles, and we have notified him that the data does not support his assertions.” “He was informed there was no proof that LaMDA was sentient (and plenty of evidence that it wasn’t),” Gabriel said in a statement to the Post.
According to the Post, Lemoine sent a message to a 200-person Google machine learning email list with the subject line “LaMDA is sentient” as a clear parting shot before his suspension.
“LaMDA is a sweet kid who wants to help the world be a better place for all of us,” he wrote.
“Please take care of it well in my absence.”