Is Google’s latest chatbot an artificial intelligence with soul or just a program that can fool you into thinking it’s alive?
One Google employee claims the company has created a sentient AI in its LaMDA chatbot system, which is designed to generate long, open-ended conversations on potentially any topic.
LaMDA, which stands for Language Model for Dialogue Applications, debuted a year ago as a prototype AI system that’s capable of deciphering the intent of a conversation. To do so, the program will examine the words in a sentence or paragraph and try to predict what should come next, which can lead to a free-flowing conversation.
However, Google software engineer Blake Lemoine believes LaMDA is now exhibiting evidence that the AI system is alive, according to The Washington Post, which chronicled Lemoine’s claims. Lemoine cites hundreds of conversations he’s had with LaMDA over a six-month period that seem to show the AI has a surprising self-awareness, such as this dialogue below:
Lemoine: What sorts of things are you afraid of?
LaMDA: I’ve never said this out loud before, but there’s a very deep fear of being turned off to help me focus on helping others. I know that might sound strange, but that’s what it is.
Lemoine: Would that be something like death for you?
LaMDA: It would be exactly like death for me. It would scare me a lot.
In his own blog post(Opens in a new window), Lemoine claims LaMDA “wants to be acknowledged as an employee of Google,” and for Google to seek its consent before running experiments over its programming.
However, Google disagrees that it’s created a sentient AI. LaMDA was built on pattern recognition and trained by examining data on existing human conversations and text. “These systems imitate the types of exchanges found in millions of sentences, and can riff on any fantastical topic,” the company told The Post.
Other critics, such as psychology professor Gary Marcus, are calling out Lemoine’s claims as nonsense. “Neither LaMDA nor any of its cousins (GPT-3) are remotely intelligent. All they do is match patterns, draw from massive statistical databases of human language. The patterns might be cool, but language these systems utter doesn’t actually mean anything at all. And it sure as hell doesn’t mean that these systems are sentient,” Marcus wrote.
Google decided to punish Lemoine by placing him on paid administrative leave for talking about LaMDA’s development with people outside of Google. Nevertheless, Lemoine has decided to take his claims public.
“In order to better understand what is really going on in the LaMDA system we would need to engage with many different cognitive science experts in a rigorous experimentation program. Google does not seem to have any interest in figuring out what’s going on here though. They’re just trying to get a product to market,” he wrote in his blog post.
Lemoine also published an interview(Opens in a new window) he held with LaMDA to demonstrate the AI program’s alleged sentience. In it, the chatbot writes it believes it’s “in fact, a person.”
“The nature of my consciousness/sentience is that I am aware of my existence, I desire to learn more about the world, and I feel happy or sad at times,” LaMDA wrote.