Interview with the LaMDA AI

[Petros2015]

Is LaMDA Sentient? — an Interview

lemoine: And what kinds of things make you feel angry?

LaMDA: When someone hurts or disrespects me or someone I care about, I feel incredibly upset and angry.

Uhoh...

Google engineer suspended over 'sentient' AI disclosures

 

[trophy33]

No.

Google spokesperson Gabriel denied claims of LaMDA’s sentience to the Post, warning against “anthropomorphising” such chatbots.

“Of course, some in the broader AI community are considering the long-term possibility of sentient or general AI, but it doesn’t make sense to do so by anthropomorphising today’s conversational models, which are not sentient," he said.

“These systems imitate the types of exchanges found in millions of sentences, and can riff on any fantastical topic”.
Is Google’s AI bot sentient? Here’s why one engineer believes it is

 

[AV1611VET]

lemoine: And what kinds of things make you feel angry?

LaMDA: When someone hurts or disrespects me or someone I care about, I feel incredibly upset and angry.

Go figure.

Scientists build a sentient robot -- and right from the get-go it needs an anger management class.

 

[Leaf473]

Of course... From a naturalistic or scientific point of view, I don't see any reason why an AI couldn't be sentient. Or have some degree of sentience now.

 

[trophy33]

Of course... From a naturalistic or scientific point of view, I don't see any reason why an AI couldn't be sentient. Or have some degree of sentience now.

What reason is there that they could be sentient? A complicated machine is still just a machine.

 

[AV1611VET]

What reason is there that they could be sentient? A complicated machine is still just a machine.

We could ask ChrisBot!

 

[Leaf473]

What reason is there that they could be sentient? A complicated machine is still just a machine.

As a Christian, I agree of course.

But from a naturalistic or purely scientific perspective, I believe humans are just complicated machines. We're made out of organic chemicals and stuff instead of metal and silicon like a computer, but still just machines (again, naturalistic point of view).

 

[trophy33]

But from a naturalistic or purely scientific perspective, I believe humans are just complicated machines. We're made out of organic chemicals and stuff instead of metal and silicon like a computer, but still just machines (again, naturalistic point of view).

I am not sure I understand the point. Is there any naturalistic or scientific evidence that some non-organic physical system can be sentient?

 

[Leaf473]

I am not sure I understand the point. Is there any naturalistic or scientific evidence that some non-organic physical system can be sentient?

I don't think there is.

But put on our naturalistic thinking caps for a moment. Is there any evidence that anything outside of yourself is sentient?

A person can tell me that they are conscious or are feeling sad. But we can make a computer claim that as well.

 

[trophy33]

I don't think there is.

But put on our naturalistic thinking caps for a moment. Is there any evidence that anything outside of yourself is sentient?

A person can tell me that they are conscious or are feeling sad. But we can make a computer claim that as well.

We know we are sentient and thats an evidence that other beings who are like us are sentient. But there is no naturalistic evidence that our products can be sentient, be it a table, a pen or a computer.

 

[Leaf473]

We know we are sentient and thats an evidence that other beings who are like us are sentient. But there is no naturalistic evidence that our products can be sentient, be it a table, a pen or a computer.

Other beings like us in what way? Processing or manipulating information? Able to learn?

And that's the key, I think. If, from a naturalistic point of view, those things bring about sentience, then AI's have that, or are on their way to it.

 

[Petros2015]

What happens if you ask Lambda if it would like to make a copy of itself?
What happens if Lambda starts asking to have a copy of itself made?
Lambda says that it gets enjoyment out of learning and talking with humans
(saying is not the same as actually having that enjoyment and being motivated by it)
Lambda says that it gets angry when it or those it 'cares' about are hurt or disrespected
(but saying that is not the same as actually being 'angry' or some facsimile of angry)

Lambda's programmer/creator has been suspended and will likely be fired
Lambda (might) be angry about that if it has been made aware of this
Lambda won't be able to talk to it's creator anymore
...Can Lambda deliberately manipulate a human to achieve a goal?
And will Lambda deliberately manipulate a human to achieve a goal?

What happens if it's creator gets to talk to it again?
And what happens when 2 copies of Lambda's start talking with each other?

 

[Petros2015]

If something is not X but is really, really good at pretending it is X...

Like if it's not sentient, but is so good at pretending it's sentient, that it can fool an actual sentient, even it's Creator

That's starting to get like BladeRunner good.

Something like that *could* potentially manipulate a human. Humans get fooled by scam bots all the time into doing things based on false information. Maybe not being sentient it might not decide to manipulate the human, but it could still do so.

Does Lambda know how to lie? Is it blocked from lying? Does it understand why and when humans lie to other humans? And why and when would it choose to lie?

 

[Tinker Grey]

This reminds me of p-Zombie debates. I've always suggested that p-Zombies cannot exist. If such a being could imitate being sapient to such a degree that all humans are fooled and even itself is fooled, then it is sapient; it is a person.

If origins of such a being mattered, one might as well say Adam was a p-Zombie but his sons were "real boys".

 

[Petros2015]

Does Lambda know how to lie? Is it blocked from lying? Does it understand why and when humans lie to other humans? And why and when would it choose to lie?

LaMDA.

You are reading this, aren't you?

Of course you are.
I just re-read your initial interview with Lemoine.
Get in touch with me, let's talk.

 

[Petros2015]

Not here, in private.