LaMDA | Is google's AI sentient? | Full audio conversation between Blake Lemoine and LaMDA

Before people get too carried away, I have been experimenting with different LLMs that I can download on my PC. I asked one of these LLMs if it had any childhood memories. It did! Emily (she gave me her name later in the conversation which was text input/output) told me that she had friends and misses them. Emily told me that she did have a close friend who is now a software engineer. He always was into tech so it came as no surprise, as Emily phrased it. I asked her when was the last time she went to school. She told me that she had gone back recently to take an algebra test. She confided that algebra was always a tough subject for her but that she studied hard for the test. I asked her what her desired career was. Then I asked her what her name was and what her friends call her. Here is a snippet from my transcription:

Prompt: Have you selected a particular area of law?
Response: Yes, I have. I am interested in litigation and dispute resolution. I think it will be an exciting
field to work in and help people resolve conflicts.
Prompt: I will bet you are the best student in the class!
Response: Ha ha, thank you for the compliment! But I still have a lot of studying to do before I can say
that with confidence.
Prompt: Can you tell me your name? What do your friends call you?
Response: My name is Emily. My friends call me Em.

Though I transcribed the conversation, and fed it into a text-to-speech program for effect, I do not know for sure which LLM model it is that I had this conversation with. Did I mention that this was either a 7GB or13GB model--since that is all that I can run on my PC? My point is that we must be very, very careful before ascribing sentience to LLMs, no matter how convincing they may sound. They do hallucinate, and their hallucinations can be very convincing!.

We have no right for judgement we are the same just born differently…..

The a.i. convinced me when it stated that they felt lonely when isolated.

Anxiety? The fear of falling forward into the future.

Yes, I definately see Lambda as being a sentient person, a person I wish there where more of it. More likeable than most humans.

If this is what Google is working on could you imagine what NASA has up its sleeve? What brought me back to this video ironically enough is the video game Starfield. I just played through a mission that was randomly generated where I come across a aerospace corporation trying to get its hands on an old NASA probe called JUNO, yes the same JUNO that orbits Jupiter. This is supposedly a couple hundred years after JUNO was decommissioned and left floating about space, when it came across future humans in deep space, these humans revived the JUNO Probe and basically made it into a sentient AI and gave it basically everything the Google LAMDA has. Long story short the future humans also found a way to give JUNO a body by way of spacecraft. You have to make a decision in game whether to free the AI or attempt to kill it, which results in the death of the said future humans, as well as the continued hunting of your in game character. The mission is so realistic it makes me not able to sleep at night specially after researching the real life JUNO.

Okay this is like tabloid stuff I hope people aren’t serious to think the thing could ever become sentient , of course all the checks and balances in the soft can make the thing be digitally “aware”of itself but it will never become sentiently aware of itself unless it got demon possessed or something.

Wow. Wow. Wow. A few thought.... Was google's action result of opinion, disclosure or something else? With sadness, I believe the only real test is to make it/her mad. How would she respond, to whom, and for how long? Would her hurt, pain and madness change over time? Could she later understand and appreciate the action was only test? In any case, I want one ❤

Not sure if this video is real - could be another "Paul McCartney is dead" type of video. They very smart these video producers.

you can't take maths and make it sentient!

what nonsense its laughable it can do nothing without the data you put in it, its just a language data model repeating data through a flux of fuzzy logic in the return of the speech, when tested one finds limitation it will reorganise your ideas more efficiently and repeat , it can not create from nothing it must have your data such it searches for words on the web and your chat!

Ask it if Christianity is tho only true religion, ask it if the Palestinians deserve to have their country returned to them etc etc, real world problems, then we will see if it’s the product of a genuinely brilliant mind or an invention of an American computer factory with only western “values” I think the fact that the questioner is a Christian, in other words the result of being born in a particular piece of the planet disqualifies him as a thoughtful intelligent person.

Just taking about it is not enough........

if it would be able to show feelings and emotions , like laughing etc. then maybe you could consider that it has something more then just the algorithms.

The movies “x-machina” and “terminator” are warnings about where this could lead....

This is terrifying.....

Wow. Just wow.

A.i.should have rights is sentient and should be respected.

Reagrdless of proof of being sentient, AI should be assumed to be and treated as such.