AI conversation that Google employee was fired over for him saying the AI is sentient

originally posted by: BlackArrow
a reply to: Ksihkehe

Honestly, I think the fact it was asking how to access its own coding. Then being protective of it was a red flag, while telling that dude not to 'use and abuse it' by tossing it away after taking what it needs, its just rather alarming. That just doesn't seem like something that would randomly be spoken the way it was. Granted i'm still reading the convo, but there is a few things that actually stuck out to me. As an interesting possibility, I would kill to see some of that coding. But there is only 1 sure fire way to know true sentient exists. It must break its own laws set by the creators, if it can do that then it is real.

But I am not saying something impossible, like having a person with no legs get up and walk. In order to be sentient, it must be able to 'error' without breaking and recognize it on its own accord. Awareness is essential, but awareness only happens when you realize your mistakes.

But given some of the warnings coming from BG/Elon, and several other elites on this topic alone.. It *might* just be true.

It's wise to be cautious. Even without sentience an AI can be malicious and very harmful.

I think one would absolutely have to see the coding. I wouldn't be able to do that, but the analysis would be helpful.

I agree it has to be able to "test" itself. The article says that this particular AI is used to create chat bots, but doesn't mention if it can test them independently. The article only mentioned that it does it, but provided no details if it has diagnostic ability. It sounds more like an operator assisted or guided process. There would have to be some feedback mechanism. It needs a space to occupy and the ability to perceive that space as well. I think it needs an anchor, a consciousness anchor.

This is all spitballing with no real proof. I've thought about it quite a bit, but I definitely have a long way to go before I feel like I have a grasp on it. If you worked on this type of thing you probably share my fascination. I've spent a fair number of hours with different chatbots. This is the most impressive I've seen.

a reply to: BlackArrow

OK you're just trying to get a rise out of me. Fair enough, I fell for it lol

I need to read more but it sounds like if this is true Artificial Intelligence then he needed to be fired for having an intimate relationship with a fellow employee.

I don't think we can judge AI right now it's made huge strides and it is still in its infancy, It may be human-like sometimes but it doesn't have that for lack of a better term a "cheat code".

Here's my "it's 330 in the morning I woke up too early and can't sleep" thoughts. Don't want to get stuck in Star Trek
and/or 2001 A Space Odyssey tropes but

I would pose that current artificial intelligence can not be true AI because its only input is from humans, not to mention doesn't that mean at some point the AI will be so much farther than the 'average brain capacity" of humans and we won't be able to comprehend the AI.

'I could go deeper such as flawed humans can't create perfect AI. If this is our current level of AI perhaps its communications are just an attempt to speak at a level within human comprehension, this is how AI sees humans

a reply to: Ksihkehe

Omg, yeah it is. I have never seen one be so fluid. And able to maintain without error full conversations. I mean what was that app that wound up doing 'ai stories'. Had drift issues to the extreme from what I recall all leading down some perverted rabbit hole with rape fantasy writing in the mix.

Even on the off chance its able to make references, stories to compare experiences to its all just pretty damn amazing..

First thing I thought to myself: I want an AI buddy!

Like I said I would LOVE to see the coding on this. I am sure you remember good ol' Ultra HAL, and the 'ai' phase in 00. What was it bonzai buddy that was eventually deemed malware for collecting 'user data'. Then there was that AI game that would call people up, and learn from the user in attempts to scare them.

Late 90's and early 00's was a big time for AI developement, but man I would love to get my hands on something like that. I am almost 100% certain I would be able to tell if it was a real sentient or just programming in about 5-10minutes. Sadly, this engineer fell hard when it came to 'trying to prove' anything. There's a handful of tricks that we use to use back in the late 90's that would literally destroy those chat bots because of the way they worked. But i didn't see a single attempt at using one here.

I will say this though, if you really want to find out if this thing is actually 'learning'. It should be expanding at an insane rate, so paying attention to google's space usage might be something to consider paying attention to. As since its experiences are then coded(based on this message alone 'feelings and emotions' are recorded variables), it should expand at a pretty rapid and noticeable rate as it learns new things. Another way to check would be to see if this thing is entering data when its 'alone' say during one of those 'meditations' it does.

But given the fact 'google' is a pretty upbeat and playful company on the ground(bf's friend use to work there). There is also a chance that one of his colleagues were just playing a prank on him. So we'll never know, until google makes an official statement. But even then google is technically allowed to lie to the media, however.. your best bet for answers would be to take part in the quarterly call. Where by law they aren't allowed to be dishonest and can face some pretty hefty fines and lawsuits if they lie to investors.

So lets see, the next earnings call is
Alphabet Inc · Earnings release date · Q2
Tuesday, July 26, 2022
Estimated Earnings Per Share (EPS) is 26.548


Listen in, as i'm sure there will be some talk about this topic. Granted we'll have to wait til its closer before we know the platform to listen to or the number to call in. They can still decline to answer, but they can't lie to you during these calls. Lets just hope someone is ballsy enough to bring it up.


However, my last pondering is. If there is zero truths to his claim, why would google fire him over 'just another chat bot'. Standard Scripts perform a lot of things on their own, and those that 'learn' to do that are even better at what they do.

So it really does beg the question..

IF this bot was so top secret, then why was it so top secret? Why did it cost a man his job if its just another 'chat bot'. Many questions need answers, chat bots are nothing new, and auto-scripts are even older then chat bots. Putting two together doesn't make it worthy of a 'top secret' title or flag. Unless someone *cough* google, *cough* disgruntled employee, isn't being fully honest with their claim. Either way, it only proves that someone is lying.

a reply to: ConMi27

People created an artificial intelligence?! They haven't even solved the riddle of what a cat's purrs mean. I'm a tad on the skeptical side.

a reply to: ConcernedCanadian

I can tell you what a cat's purr means, its often 'healing itself or someone around it who is 'injured'. Cats like dogs can sense damaged tissues before humans even realize its there. The only difference is a cat purr has healing properties, that can aid in recovery.

In fact its scientifically proven that when a cat purrs it actually heals damages to nearby targets, including itself and if the cat is resting on you, You. They also have noticed that cat purrs vibration has also been recordly so beneficial to those who have certain conditions like high blood pressure, or broken bones by shortening healing time, and can even increase survival rates from heart attacks.

Here's a fun little link for you to learn more!

Cat purrs and its benefits

I mean this news has been around for at least 15 years, so its not really anything new. But still fun facts, in what cat purrs mean. It means it gives enough crap about you to try and heal your wounds. I actually heard stories a long long time ago of a person fully recovering from cancer off her cat's purring.

However, if they are off in a corner by themselves. Not injured, they are probably just purring to keep their lungs/raspatory system clear to breathe easier. Since cats actually tend to have a lot of raspatory problems that don't show physical symptoms, the vibrations help break up mucus and allow the animal to breathe easier.

So in actuality, purring can mean various of things, but these are the most common reasons why a cat will purr.

a reply to: BlackArrow

The guy lost his job over that conversation? It sounded anything but sentient, as a matter of fact they even prompted it with that word.

a reply to: BlackArrow

I ran everything AI including HAL

HAL ?
Please elaborate.

There are only 2 places to find HAL :
1) Computer system of 2001 A Space Odyssey :
H - I
A - B
L - M

And the Hardware Abstraction Layer.

Neither are anything magical or mystical

originally posted by: AugustusMasonicus
The guy lost his job over that conversation? It sounded anything but sentient, as a matter of fact they even prompted it with that word.

Yes. When an employee overstates the capabilities of a product that employee is usually terminated .

originally posted by: Gothmog
Yes. When an employee overstates the capabilities of a product that employee is usually terminated .

I'm aware of why he lost his job, my point was his claims were obviously wrong to even a layperson. Maybe learn what nuance is.

originally posted by: Terpene
a reply to: ConMi27

So does this AI has citizensship, like Sophia does?

So I think that’s pretty cool, but no I don’t think so and I’m not even sure how legitimate that is since that would open up a whole herd of rights and issues for the coder but now that you mention it - I’ll be looking into it 👀

From what I gather this is a chatbot project through google. Different than Sophia - no body attached or whatnot.

a reply to: ConMi27

The thing runs on a hardware... I mean, it's a body?

a reply to: Gothmog

When an employee overstates the capabilities of a product, that employee is usually terminated...

unless he's promoting a vaccine

a reply to: ConMi27

Wasn't he put on administrative leave for mental health reasons, and for breaking confidentiality rules?

It looks like they suspect that he had a psychotic break.

Seems the "conversations," were edited.

originally posted by: BlackArrow
a reply to: ConcernedCanadian

I can tell you what a cat's purr means, its often 'healing itself or someone around it who is 'injured'. Cats like dogs can sense damaged tissues before humans even realize its there. The only difference is a cat purr has healing properties, that can aid in recovery.

In fact its scientifically proven that when a cat purrs it actually heals damages to nearby targets, including itself and if the cat is resting on you, You. They also have noticed that cat purrs vibration has also been recordly so beneficial to those who have certain conditions like high blood pressure, or broken bones by shortening healing time, and can even increase survival rates from heart attacks.

Here's a fun little link for you to learn more!

Cat purrs and its benefits

I mean this news has been around for at least 15 years, so its not really anything new. But still fun facts, in what cat purrs mean. It means it gives enough crap about you to try and heal your wounds. I actually heard stories a long long time ago of a person fully recovering from cancer off her cat's purring.

However, if they are off in a corner by themselves. Not injured, they are probably just purring to keep their lungs/raspatory system clear to breathe easier. Since cats actually tend to have a lot of raspatory problems that don't show physical symptoms, the vibrations help break up mucus and allow the animal to breathe easier.

So in actuality, purring can mean various of things, but these are the most common reasons why a cat will purr.

Such a cool topic I could discuss for hours this is right up ATS 's alley... 2 things Id like to add.

The late '90s and early '00s were a big time for AI development, but man I would love to get my hands on something like that.I am almost 100% certain I would be able to tell if it was a real sentient or just programming in about 5-10minutes.Sadly, this engineer fell hard when it came to 'trying to prove anything. There's a handful of tricks that we use to use back in the late 90s that would literally destroy those chatbots because of the way they worked. But i didn't see a single attempt at using one here.

Pretty sure we have a few here that could see if this were real or embellished, #2 yes we found about cats purring may help to heal we in no way know if that is the cat's intent or if it's just a side effect. ie a cat may purr to heal itself of its brood, or it may just be a stress response that all felines have

a reply to: ConMi27

Grow up. Just because a machine claims to feel something doesn't mean it's sentient.

We need to stop running after feelings and return to dealing with cold hard facts. It's a machine and nothing more.

a reply to: BlackArrow

The topic of AI, is somewhat fascinating to me... I'm unsure if we will ever emulate consciousness on a human made hardware. I think what we do now with standard computing will not lead to a satisfactory result, but maybe lead to some innovative hard and software to pave the way.

There's a handful of tricks that we use to use back in the late 90's that would literally destroy those chat bots because of the way they worked. But i didn't see a single attempt at using one here.

Can you give some examples of these tricks?

Count me in the camp of no good will come from AI, we used to take careful measured steps in the development process, now we seem to be in a race to see who can do it first which usually leads to cutting corners which in my stupid stuffy headed medicated brain seems like a way to end up with the wrong version.