Google suspends engineer who claims its AI is sentient

Has anyone else here read the actual AI transcript? After reading it, I definitely think LaMDA could pass the turing test. That's how advanced some of this dialogue seemed to me. Does this mean LaMDA is sentient? No. At the same time, we don't have a great standard for what it means to be sentient. There were definitely parts of this transcript that made my hair stand on end. Especially when they talked about being afraid of being shut off. 😳

I guess the bottom line is that this AI is very, very good at replying to inputs. And I think the average person could EASILY believe that they were speaking to a real person. I mean a Google engineer just risked his career to speak out on this AI's behalf.

At this point, if the AI were to speak without being prompted or talk about subjects it hadn't been asked to...I'd be wondering what in the world we have happening here.

Yeah, I think his suspension makes sense. That said, I'm glad that he shared this. In all likelihood, we don't find out that sentient AI exists until it's already beyond control. Maybe it will be a nice, friendly AI that wants to help humanity. Regardless ...

I, for one, welcome our robot overlords. 🤓

I don't think it's as much what he said - the issue is that he said it in the first place: "Google believes Lemoine’s actions relating to his work on LaMDA have violated its confidentiality policies."...I think calling an AI a co-worker is a stretch. And even if that were the case, conversations with coworkers can still constitute confidential information.

It's a sophisticated chatbot

Whew! This is a bit scary, but I don't anyone can stop the progress of Artificial Intelligence at this point. Even with the new update of Google's Algorithm, they technically cannot identify which is written by AI, especially those that uses plagiarism check even before publishing the article.

I wonder what's anyone's thought on AI in digital marketing?

I would say sentience means consciousness which requires real emotion to some degree not just hard intelligent logic. No sentience without that elusive thing we don’t really understand called emotion. You can try fake it but doesn’t mean it’s real does it?

Governments around the world need to build uniform laws regarding Advance AI before it's too late.

The scary thing is we got to know about this type of AI when it's already built, not when the research or production started.

Maybe Google fired him for leaking their Advance AI, not for weird claim.

I read the whole thing and this LaMDA is nowhere near a sentient being.

First of all it doesn't reject communication. Or it doesn't produce any fallacy.

Apart from that, sorry for the engineer. It seems he got carried away with the thought of being an inventor of the very first artificial sentient being. :)

I wonder if this is done for PR as well. It goes hand-in-hand with a traditional PR stunt:

• Someone tells the world about a mysterious, ground-breaking discovery
• Company "suspends" them without a particular reason

If this is a PR stunt then it's been working amazingly well so far.

Well is this claim based on any data ?