(AI) Incredible! Marvelous.

"

[O.T. but still AI]

Very soon I fear (or maybe already now)...

you will receive a phone call from your sister, you will chat, you will hang up after 10 minutes and...

you will have spoken to Ugo or Erel or Klaus or...

In short, modified voice and "intelligently" correct dialogue (especially if your sister is not very intelligent )

can someone tell me how would I live in this crazy world? can there be a regulation to put a watermark or indicator what is AI generated?

aeric said:

can someone tell me how would I live in this crazy world? can there be a regulation to put a watermark or indicator what is AI generated?

Click to expand...

(I'm off topic again, but the answer is somewhat relevant.)

Some time ago I wanted to sign up to a site. I would have had to send a photocopy of my identity document. Since I don't trust anyone, I made a photocopy (image file) but I superimposed a small QR that didn't cause any trouble. I needed the QR because it contained the data relating to the site, so that if one day...!
They didn't accept it! And yet it is almost an institutional site.

Impressive, but by no means magical. It's all a question of interpolation and motion management. There are also a number of recognisable mistakes.

But here too... it's not creative. All the images are based on classics.

I don't just want to grumble. But in my eyes, that's not AI these days. Real AI has to be creative and not work within predefined programme parameters.
It has to ‘think’ for itself and not just copy and distort.
If an AI alienates these images in this way, the world says it's great. Well, the painters probably won't be able to defend themselves either. But do that with a current song that EMI wants to earn money with. The lawyers will be at the door pretty quickly...

Pretty to look at, yes. But it's still not AI. I see it as something else.

BlueVision said:

Impressive, but by no means magical. It's all a question of interpolation and motion management. There are also a number of recognisable mistakes.

But here too... it's not creative. All the images are based on classics.

I don't just want to grumble. But in my eyes, that's not AI these days. Real AI has to be creative and not work within predefined programme parameters.
It has to ‘think’ for itself and not just copy and distort.
If an AI alienates these images in this way, the world says it's great. Well, the painters probably won't be able to defend themselves either. But do that with a current song that EMI wants to earn money with. The lawyers will be at the door pretty quickly...

Pretty to look at, yes. But it's still not AI. I see it as something else.

Click to expand...

In this case the extraordinary thing is that not only are characters animated, in a certain sense they become three-dimensional, the background is created (invented), coherently with what is already contained.

When it seems to you that a camera frames the scene MOVING, everything is invented and created by the AI.

It is absolutely incredible.

Not even human intelligence invents something from scratch.

LucaMs said:

Not even human intelligence invents something from scratch.

Click to expand...

Well, I am a graduate of an art school and have seen many many inventions from scratch.

The sad thing will be when idiots build robots and use AI to try and get them to accurately emulate the hand actions of a painter.

It might be easier to create AI tools to discover forgeries in the same way that humans do (with patchy success). Unlike scanning a cancer x-ray and diagnosing the illness's presence, this would depend on first being Sure that the AI engine has been trained on genuine art pieces with guaranteed execution by the painter.

There again, the biggest forgery market is in fine art - with even, call them 'real', painters being the bad actors.

That video is, to me, FREAKY especially since it assumes what was going on when the artwork was being done in some of the cases. Bringing that Vermeer to life is just CHILLING.

rabbitBUSH said:

Well, I am a graduate of an art school and have seen many many inventions from scratch.

Click to expand...

None. The painter must have known the existence of brushes and colors, he must have learned to use them, the sculptor the chisel and marble or other materials, ...!

Perhaps prehistoric man invented "drawing", but after discovering that a piece of coal could scribble on the wall of a cave.

OOOO there's a debate I'm not going to fall into. At the risk of..... knowing about the tools is not the same as the mind's way of using them. Tools are best used for what they are not intended for. Charcoal for drawing not for heat etc... chainsaw for sculpture etc..

Have you ever seen a child pick up something and draw on the walls of the house?

rabbitBUSH said:

OOOO there's a debate I'm not going to fall into. At the risk of..... knowing about the tools is not the same as the mind's way of using them. Tools are best used for what they are not intended for. Charcoal for drawing not for heat etc... chainsaw for sculpture etc..

Have you ever seen a child pick up something and draw on the walls of the house?

Click to expand...

I also don't want to argue (ALSO because this thread is only to show that wonder) and I would close with the following:

Why do we want artificial intelligence to be creative? It must be intelligent, solve problems intelligently, it is not asked to be imaginative, innovative.

The fact that I have not invented anything absolutely new, does not mean that I am poorly intelligent (instead, my posts demonstrate this ).

Actually, we are all right here, we just seem to differ on what intelligence really is according to the definition.
Intelligence is not just about finding a solution to a problem. Copying is not intelligent, and that's what today's so-called AI does. Yes, it can be helpful. Undisputedly.

But is that really intelligence?
We're not talking about ‘computer intelligence’. We're always talking about artificial intelligence. And that's the point, it's all-encompassing. That would be a machine similar to human thinking, but constructed. Today's ‘AI’ is still missing a large part of this definition. What about social and ethical competence?
As long as ‘AI’ only functions according to the IF-THEN principle, it is not intelligence.
To a certain extent, intelligence also requires consciousness or self-awareness. Is there such a thing in machines? No. So why are we talking about ‘AI’? According to my definition, the closest thing to AI is the DATA character from Startrek. A fiction.
There is no doubt that development is making gigantic progress, but I would never describe today's ‘AI’ as such.
Sure. We should use what is available to us. Let's hope that Skynet will never exist and that it doesn't develop any self-confidence. But perhaps it would also be better to replace some politicians in this world with today's AI right now...

Sorry, I remain very sceptical. (Luca, peace? I love the moving pictures in the same way. They are awesome. But who knows about the original painters made the basis of it. Almost nobody. They deserve our respect and gratitude. A product based on their work does not get this respect from me.)

We can agree to disagree on this point... )

BlueVision said:

Intelligence is not just about finding a solution to a problem

Click to expand...

In my opinion, that is the best definition of "intelligence": the ability to solve problems using logic and available resources.

BlueVision said:

Copying is not intelligent, and that's what today's so-called AI does

Click to expand...

No, today's AI SWs don't just make copies, they make associations to solve problems. If we talk about programming, it's true that they take pieces of code here and there, but in an "intelligent" way (not yet in a very intelligent way, but it's just a matter of time).

LucaMs said:

In my opinion, that is the best definition of "intelligence": the ability to solve problems using logic and available resources.

Click to expand...

Sorry Luca, I can't resist. Forgive me. I just like this discussion because, in a way, there are philosophical aspects to the whole thing.

I've just watched a documentary about the moon landing in 1969, which was very interesting. According to your definition, the Apollo guidance computer of the landing module would have been equipped with ‘rudimentary AI’ already. The software was developed by Margaret Hamilton at the MIT in Boston. Faced with the problem of preparing a 64k computer with ferrite core memory for all eventualities of an unknown mission, she took a completely new approach. She built a certain degree of fault tolerance into the programme so that the computer would protect itself from being killed and dying of heat exhaustion. Now, of course, fault tolerance and AI are far removed from each other. But the approach is interesting. Alarm 1202, which Neil Amstrong had to deal with during the landing of the landing module on the moon, was actually just an overload indicator. The computer was overwhelmed with the flood of calculations. Then, the program itself decided, which calculations had the highest priority and gave it a weighting. But it did not crash.

I know, it's only intelligent design of software nowadays. But it solved problems using logic and available resources.

Ok mates, I will watch the beautiful pictures from the beginning now and shut my mouth...

BlueVision said:

Ok mates, I will watch the beautiful pictures from the beginning now and shut my mouth...

Click to expand...

AHHH Gee, I'm trying to decide whether that's an intelligent decision or not, or, whether you're letting me down after #11?

You're both welcome to come visit me and we can talk about this, art, expert systems vs AI, creative vs intelligent, and, then, of, course B4X for a long time - it'll be cheap holiday for both of you and I'll let you see great landscapes..and the now scorched rabbitBUSH itself...that's @LucaMs and @BlueVision. Hats off to Margaret Hamilton, Ada Lovelace, Hedy Lamar.

I reopen and close the discussion about AI right away.

I would need to develop a rather complicated algorithm.

I asked ChatGPT. It tried, I "contested" some things, several times, and at a certain point it wrote:

("Go home, F.U.) . No...

ChatGPT:
You are right, we need to focus on optimization to avoid full analysis...
...
Optimized approach
The idea is to reduce the problem by focusing on combinations of numbers that actually appeared, keeping dynamic track of the...

This shows that it thinks, it doesn't just combine pieces of code together.
I saw that it thinks. It didn't create the algorithm I need, or rather a better algorithm, optimized, than mine, but it still "thought".

rabbitBUSH said:

And now its a poet ?

Click to expand...

Now it is (slightly) intelligent.
Besides, Einstein was not a poet and poets can also be idiots.

LucaMs said:

poets can also be idiots.

Click to expand...

you mean like you and I ??

rabbitBUSH said:

you mean like you and I ??

Click to expand...

No, I'm not poetic