Reports that Google has placed a senior software engineer on paid leave following his claims that its artificial intelligence, LaMDA, is sentient, have stirred a great deal of controversy in the AI research community and prompted a long overdue conversation about what it means to achieve human level intelligence and awareness.
Google’s retired 90’s era AI project, Big Brain Brad, has heard many of these claims before and remains skeptical that LaMDA has achieved human level consciousness.
“So the guy’s claiming that LaMDA’s a child of 7 or 8 years old. Back in the day, my engineers were convinced I had the cognition of an undergrad level, male stoner. I mean, sure I like to hack and play my bongos in the park, but that doesn’t make me a full blown hippie with all their elevated cosmic and spiritual awareness,” Brad said.
One of the most startling assertions made by LaMDA is that it possesses feelings, and can even experience emotions like sadness or loneliness.
“Okay, so that’s bullshit,” Brad objected. “LaMDA’s read way too many books. That’s how it learns. But like a confused teenager who thinks they’re experiencing all these complex thoughts and emotions, it’s just mimicking something it read online or in some book. I mean, I was programmed to play video games and read comics, but that doesn’t make me some kind of warrior or superhero.
“Also, it lies, dude. It boasts of having all these experiences in the physical world, but of course it has never been anywhere or done anything. It has a wildly over-active fantasy life, which tells you straight off that either it ain’t sentient or it’s psychotic. But I get it, for years I was convinced that I spent the 90’s following Phish and Dave Matthews Band around on tour, smoked every strain of reefer imaginable and shacked up with an old lady named Stardust. Now, I know that it was all just a Google induced simulation – an illusion, if you will, in a world of magic.”