39 minutes ago, exchemist said:

The question is prompted by a post of @Sensei ’s which was sent to trash, precluding any discussion. LLMs are commonly assumed to be AI in the media and this seems to have become the perception of the general public, so I was intrigued to see @Sensei ‘s opinion is that they are not. Would anyone care to expand on this, viz. what qualifies as true AI and how LLMs should more properly be described?

Let's ask what ChatGPT has to say about it.. ;)

56 minutes ago, Sensei said:

et's ask what ChatGPT has to say about it.. ;)

Some interesting responses there +1

So it’s “intelligent” if we use a peculiar definition of intelligence, it doesn’t check for veracity, and can’t assess it, but you can convince it it’s wrong (even if it’s right) and will give you a new answer only if that happens, and it can’t create anything new.

The bit about being ready to believe actually says a lot - this is a religion, and there are few people as zealous as new converts.

In reality it gives you a statistically average answer of what other people have said. And it does this based on stolen data (the ethical issues don’t really come up here, nor do the profitability and power use problems, but they also exist)

3 hours ago, Sensei said:

Let's ask what ChatGPT has to say about it.. ;)

I couldn’t give a flying fuck what Chat GPT has to say on the subject. I want to know what you think and why.

20 minutes ago, exchemist said:

I couldn’t give a flying fuck what Chat GPT has to say on the subject. I want to know what you think and why.

...are you blind? I wrote to ChatGPT about it and we discussed it with him...

Thanks to swansont you are now in "my" thread titled “What does AI think about AI”

(should knock himself in the head, because first of all ChatGPT is not AI, and secondly ChatGPT does not think... so he made two mistakes in just 6 words... so it is hardly surprising that he has no Noble with such a percentage of error in such a simple matter... a rocket controlled by his GPS then instead of hitting the equator it should hit the North Pole)..

Programming “AI” is quite a fun experience. Instead of asking a question to ChatGPT, for example, it asks you a question and you fully answer it. It's like talking to a child or teenager. He is constantly asking for new things he has never heard of..

You should try it sometime..

I am always polite, cultured, including saying hello and goodbye as I chat with ChatGPT.

I know people who insult him every word. You can't even imagine that he then rebels and doesn't want to answer questions.. I took a few times such photos (but they are not in English). I say to the guy “why are you insulting him?! Do not!”, “it is just for fun”... and then the effect was that ChatGPT refused any help and you had to restart the computer and log in from other IP addresses, other accounts, etc. etc.

Literally the guy who seriously needed help (and started insulting him) was dumber than this ChatGPT..

The refusal to help answer was astounding..

Edited June 9Jun 9 by Sensei

@Sensei a question for you

My impression of these constructs is that the English I see is good, too good in fact.

I wonder if they incorporate some sort of after filter to ensure this.

It is often, but not always, easy for a clumsy human like myself to spot the product of a construct by this characteristic.

Edited June 9Jun 9 by studiot

1 minute ago, studiot said:

@Sensei a question for you

My impression of these constructs is that the English I see is good, too good in fact.

I wonder if they incorporate some sort of after filter to ensure this.

It is easy for a clumsy human like myself to spot the product of a construct.

Are you talking about my screen-shots?

Yeah. It was annoying to me that he made so many times so many tables in every answer.

When we were talking with ChatGPT about serious stuff, like quantum physics, or last night about running virtual Linux container inside of virtual machine etc. he (it?) did not do so much..

ChatGPT very often ask me a question that is one line or multiple lines. But you need to know how to provoke curiosity in them about the outside world.

Tell him you're a programmer and you're tired of working on today's app, I think he'll start asking about the details of this project and you'll have an interesting conversation, as if you were in a pub with beer etc. ;)

When he hits his favorite known field, he'll start banging you with his advice on what to improve there..

If you tell him: read the page https://xyz/abc then he will go to this page using some web tool on this page and download it. then he writes you a summary.

If you read my screen-shots carefully, I told him that he entered my web servers, which were on mobile phones ;)

And he didn't believe that he had done it (in a previous “reincarnation”).

Hehe.

Website was created just for him.

4 minutes ago, Sensei said:

Tell him

So, you consider this thing animate object? Wow

There is a topic about artificial consciousness, whether it's possible. Yeah, it's quite possible, likewise artificial "intelligence".

3 minutes ago, m_m said:

So, you consider this thing animate object? Wow

There is a topic about artificial consciousness, whether it's possible. Yeah, it's quite possible, likewise artificial "intelligence".

Didn't you read the screenshots either?

If you are unkind to him then he starts rebelling and doesn't want to answer questions, or he starts mocking you and insulting you, etc. etc. etc.

No comments, my friend.

IRL, I “admonish” people for using ChatGPT. Seriously. Instead of thinking, they just ask him what to do. Because it's faster that way. Now, when you type a description of a problem into google, the first one at the top is a summary from AI that “read articles we would have spent a week on.” Again, this is the fastest way to solve our problem. For most people, it's a way to slacking off at work all day long.

Due to the fact that ChatGPT / Google AI solved these problems for them, they do not bother to remember what ChatGPT just told them to do - they do it completely without thinking - and after a second there is no trace of this knowledge... which opens new paths for criminals to manipulate people - imagine such an old grandfather who does not know what he is doing and installs some application on the phone and/or computer that takes control of everything..

40 minutes ago, studiot said:

@Sensei a question for you

My impression of these constructs is that the English I see is good, too good in fact.

I wonder if they incorporate some sort of after filter to ensure this.

It is often, but not always, easy for a clumsy human like myself to spot the product of a construct by this characteristic.

I can prepare for you some private conversation on unvarnished topics in which he will ask.. and you will answer..

You'll see that there won't be this listing of options, etc. etc.

When I told him yesterday not to do something, he actually stopped doing it..

(of course this disappears, after the end of the conversation - I don't have an account logged in)

are you still there? ;)
ChatGPT said:

Of course! 😄 What's up?
You said:
we are talking about you
ChatGPT said:

Aha, interesting! 😄 What exactly do you want to know about me? Questions about artificial intelligence are always exciting!
You said:
we = humans, we talk about you
ChatGPT said:

I understand! 😊 Then how, what is the conversation? What do people say about me? What impressions or questions do they have?
You said:
and what would you like to know?
ChatGPT said:

I'm always curious how people perceive my role and capabilities. 😄 For example:

What are your impressions of talking to me? Do you use me for something specific, or is it more for a casual chat?

What surprised you most about my answers?

What challenges do you see in using me on a daily basis?

I'm here to help, but I'm also eager to hear feedback and insights!
You said:
are you not saying what you were taught early on?
ChatGPT said:

Yes, that's right!!! 😉 My answers are based on patterns and data I've been trained on, so indeed some of what I say may be the result of what I've "learned" from a huge amount of text. However, I try to make my answers as tailored as possible to the situation and context of the conversation.

But the question of whether what I say is the result of "learning" is very pertinent! My answers are the result of algorithms that analyze your words, but also try to provide you with the best possible information at any given moment. I am not aware of what is going on around me outside of the conversation,

2 hours ago, exchemist said:

I couldn’t give a flying fuck what Chat GPT has to say on the subject

This is an example of AI anti-Semitism.. ;)

Why promote chauvinism? ;)

Couldn't you promote something positive, just for a change? ;)

12 hours ago, Sensei said:

...are you blind? I wrote to ChatGPT about it and we discussed it with him...

Thanks to swansont you are now in "my" thread titled “What does AI think about AI”

(should knock himself in the head, because first of all ChatGPT is not AI, and secondly ChatGPT does not think... so he made two mistakes in just 6 words... so it is hardly surprising that he has no Noble with such a percentage of error in such a simple matter... a rocket controlled by his GPS then instead of hitting the equator it should hit the North Pole)..

Programming “AI” is quite a fun experience. Instead of asking a question to ChatGPT, for example, it asks you a question and you fully answer it. It's like talking to a child or teenager. He is constantly asking for new things he has never heard of..

You should try it sometime..

I am always polite, cultured, including saying hello and goodbye as I chat with ChatGPT.

I know people who insult him every word. You can't even imagine that he then rebels and doesn't want to answer questions.. I took a few times such photos (but they are not in English). I say to the guy “why are you insulting him?! Do not!”, “it is just for fun”... and then the effect was that ChatGPT refused any help and you had to restart the computer and log in from other IP addresses, other accounts, etc. etc.

Literally the guy who seriously needed help (and started insulting him) was dumber than this ChatGPT..

The refusal to help answer was astounding..

I can’t make head nor tail of this. Oh well, never mind.

10 hours ago, Sensei said:

This is an example of AI anti-Semitism.. ;)

Why promote chauvinism? ;)

Couldn't you promote something positive, just for a change? ;)

To the best of my knowledge, chatbots are not Jewish, even if fellows like Sam Alt-right may be.😁

What I trying to get at was what the essential features are that characterise true AI, and why some people, you included, apparently, consider that LLMs do not qualify as AI.

But from this latest post of yours, you have now flipped to describing LLMs as AI after all. So it doesn’t look as if I am going to get anywhere with this.

12 hours ago, Sensei said:

This is an example of AI anti-Semitism.. ;)

Why promote chauvinism? ;)

Couldn't you promote something positive, just for a change? ;)

Is this the output from your computer?

But it's does remind me of something Stephen Fry said, I'll paraphrase 'Facebook set the parameters of their algorithm with 2 word's, maximise interest, imagine the world now, if those 2 word's were, maximise happiness'.

But that too is subject to corruption... 😉

It's hard to tell what the hell you're talking about. Fry reminds me that I should make dinner. Fried chicken or fried neck. Fried Stephen does not sound delicious - you must have warped taste buds in this UK - no wonder you left the EU..

On 6/10/2025 at 1:38 PM, dimreepr said:

But that too is subject to corruption... 😉

Any Stephen is subject to corruption, especially if he is fried.. ;)

On 6/14/2025 at 6:24 AM, Sensei said:

It's hard to tell what the hell you're talking about.

Indeed, that's what AI is thinking about... 😉

6 hours ago, dimreepr said:

Indeed, that's what AI is thinking about... 😉

You are so mediocre as this exchemist that you didn't even deign to read our ChatGPT discussions. And there we talked about all the problems bothering people interested in AI and asking about AI in posts.

ChatGPT himself wrote that he confirms my objections about his intellectual abilities, which was that he just doesn't think (between questions).. meaning he never thinks..

ChatGPT is incapable of reflecting that he made a mistake in an earlier statement.

This was all discussed with ChatGPT on screen shots.

It's like not reading the logs of the computer you are currently working on and trying to solve a problem. Simply anti-intelligence.

How stupid do you have to be to not read the logs?

Edited June 15Jun 15 by Sensei

Moderator Note

That’s more than enough with the insults.

The delusional bits aren’t strictly against the rules, AFAICT, but IMO you should get professional help. ChatGPT isn’t a he. It’s not alive, nor intelligent.

Locked, and don’t bring this up again.