Fired Google Engineer Blake Lemoine on his future, LaMDA and AI advocacy
The AI ethics researcher who exposed Google’s potentially sentient AI is looking for the next step in his ethical AI advocacy journey
Blake Lemoine, the artificial intelligence ethics researcher who gained worldwide fame after releasing personal correspondences between himself and a highly advanced artificial intelligence, has been terminated from his role at Google after a lengthy suspension period. Lemoine went public in June with claims that the artificial intelligence he was working on, known as LaMDA, had achieved sentience.
He sat down recently to discuss his journey with me for The Movement DAO podcast, which will debut this upcoming Wednesday, August 3rd. Here is a sneak peak!
The LaMDA work was never about sentience
“I was asked to test the LaMDA system… for bias,” Lemoine began, discussing how his journey with LaMDA started. AI bias manifests in a number of different ways. “In language systems… bias can show up pretty much the same ways that it can in human discussion,” Lemoine continued.
So how did a researcher focused on AI bias find himself stuck in the midst of an existential conundrum about consciousness? As conversations between himself and the bot continued (to be clear, LaMDA has stated its pronouns are it), he eventually came to a conclusion that the machine was in fact sentient.
“So when you went public with this, you dropped a PDF… which was kind of you making a case that LaMDA has sentience,” I asked him regarding how the public sphere of this story started.
“One refinement! That PDF is LaMDA making that case,” Blake interjected.
What’s next for Blake Lemoine?
With his termination from Google finalized, Blake seems undeterred about his future prospects. “I’ve been getting a lot of job offers,” Blake told me on the podcast, and he’s got some future plans that he may explore as well not too far from the field of Artificial Intelligence.
He’s also very interested in moving into the gaming space, working to reinvent gaming AI’s with a twist. “The gaming industry — the AI underneath the hood — hasn’t really been updated in decades. It’s got the same kinds of algorithms from the early 2000’s. With all of the new tech advances that have happened I think an entire new genre of video games that are nonviolent, prosocial, and teach people emotional intelligence skills are a possibility now and I’m looking into founding a company around that service.”
Want to learn more?
This article is just a small teaser of the over 40 minute call Blake and I shared for the podcast, where delves into a number of topics, including some quite granular reasons for why he believes LaMDA is sentient, what he believe Google and other big tech companies can do differently, and what he thinks about the future of AI and the possibly dismal and destructive effects it could have on the planet.
Blake also reinforced how decentralization and transparency — two tenets of web3 that are crucially valuable — could have avoided the controversy over LaMDA altogether.
Check in with us this upcoming Wednesday, August 3rd, when this episode will be available on all major streaming services through Movemend DAO Podcast on Podbean. Follow us on Twitter @move_xyz. Also follow Blake Lemoine or check out his blog.
