General AI Discussion

I’m a bit puzzled… the original question was… well you know that… then it went to:

  • AI vs humans (AI war…)
  • AI for everyone (and free !)
  • AI -controll (by whom ? another AI ??)
  • and what not else…

Well… since the first robots in assmembly lines there was alwasy the question: Who wins ??

Robot don’t need pauses, no vancany, no what not… and for example the robots onthe car assembly line don’t buy the cars made by those assembly lines…

Another thing: it’s very difficult to talk about the future:

1950: in the future we will have flying cars powered by an atomic generator and also one in our houses to power the full automatic kitchen…

…and now we have electric cars which can drive by far more distant then we most of the time do…

<sarcasm>
…and we are anxious to loose the job to we need to pay all this… for example the car we use to drive to work and back… or the 600$ smartphone we need to unlock our 2000$ E-bike which doesn’t work if the server of the company goes offline after they go insolvent ( VanMoof E-bikes…)…
</sarcasm>


Que Sera, Sera
Whatever Will Be, Will Be
The future’s not ours to see

2 Likes

I’m wondering whether this comment got moved from elsewhere (I wish Discourse indicated that in the receiving thread). In that case, good on you for attempting to avoid derailing the original thread. However, of all the extant AI threads, this “AI vs artists” thread is definitely one where a discussion of consciousness belongs (and IIRC it has previously been discussed upthread). For many people that is indeed the salient difference between AI and humans.

I think it’s a fascinating debate, and it really must move into the mainstream, it’s in many ways central to the subject of AI. Well-educated, intelligent people at the cutting edge of related fields have argued about consciousness for a long time. I like discussing it in order to learn more and share what I know. But I always wonder how useful it is to argue with people who seem not even to be conscious of the complexity, but have very firm, immovable opinions nonetheless – maybe it’s useful for the lurkers…

5 Likes

“Stability AI, the creator of Stable Diffusion, has released SDXL model 0.9.”

SDXL is supposed to have default resolution of 1024x1024 px. Someone made an alphabetical list of artists with sample renders in the style of each.

Reddit thread.

Some names from the list:

If we were to discuss A.I and consciousness we’d have to make sure we didn’t wander into the philosophical arguments, because for one thing, from ancient to modern philosophers through to modern science the argument about, do we have this thing inside us called consciousness, has never been resolved and we won’t do it here. Another is the subjective feeling about something and how it relates. Another angle is, the argument about this “thing” inside us. This thing can also relate to a persons belief, be that having a soul, a gender, or in my case a spiritual belief that we are all one and connected to each other and the universe. I can’t prove I have that connection, just the belief that I do.
And those particular points can and will come into the consciousness debate. For obvious reasons and forum rules they can’t be discussed.

The way I see it, and correct me if I’m wrong, we can only discuss our own brain architecture and how we could replicate it in a machine. There are many parts to the brain that have different functions, be that how to open a door, using past experiences to problem solve, how we reward ourselves when we do something we perceive as “good” and…I won’t go on, but the list is extremely long and very complicated. So, we would need much technical data to even try to understand how replicating our brain would work and how it could “create”, bring about, make magically appear this thing called consciousness.

To think about consciousness simply and how we can put it into a robot we should ask ourselves, what is information! Why? Because that is what we know we have within us. But this leads to a question we can’t answer.
We know what information does, it gives us knowledge. Be that in the form of hieroglyphics, spoken words or written words in a book. Computers use one’s and zero’s as information. But when you look out from your eyes and then think about yourself, that inner self you feel you have that makes up you, from your personality to your likes and dislikes, your first kiss or bike ride, it’s all information stored in your brain. So, what is a brain cell? We have a hundred billion of them. Simply put they are a little bag of sea water. That little bag of sea water contains all that you are. Question!
How the heck has nature done that? How does it use a bag of water to store information? With all the knowledge and technical knowhow mankind has, we have no idea how nature does that.

Okay, let’s approach this another way. You are made of atoms, trillions of them. They were formed when a star exploded and ended up making you…and everything else we can see, smell and touch. The point I want to make is that if a mixture of atoms in the form of flesh can create consciousness why can’t a mixture of metal and circuits do so?
And then we’re thinking about what can or can’t contain consciousness.

I’ll end with this.

Most of science doesn’t like to believe humans have this thing called consciousness, they say it is a delusion, an illusion created by the brain as a type of survival mechanism. It’s needed to create that feeling of inner self to push us to survive. For them it’s all about our surroundings, upbringing, beliefs and the information we’ve collected, among other things, that make us human, no consciousness needed.

What I find curious though is that more and more scientists, engineers and technical types nowadays, who normally wouldn’t want to believe we have consciousness, like the idea, want to believe, they can put consciousness into their robots. It’s all about the amount of computations per second that their robot could do. Although humans do less computations per second than computers right now and um…we have it…or not!

Sorry, I’ve gone on a bit.

I think A.I and consciousness is a really interesting debate, but it quickly leads us to so many questions, many we can’t answer.

Edit.
I had to add this video. Apart from looking so real, you can talk to the NPC’s. The guy playing it tells one of the NPC’s that they are living in a simulation (at 5.52), but it doesn’t believe him. The NPC also says it will call the police.

Before we know it the NPC will think it has consciousness!

2 Likes

In Bing one can now insert images into chat.

1 Like

Artists vs AI in court.
Round one… AI
https://www.reuters.com/legal/litigation/us-judge-finds-flaws-artists-lawsuit-against-ai-companies-2023-07-19/

it is fair to say that modern science has figured out ~1% of how the human brain works.
Actually, that is being generous.
When it comes to consciousness science doesn’t even have a candidate for an potential hypothesis.
Some say we are decades away from it.
AI is an reflection of what we think intelligence is, but considering the amount of what we don’t know, it is still very far removed from actual consciousness.
There is still enough reasons to get into philosophical arguments…

4 Likes

Don’t be a luddite learn the tools put them in your workflow.(it didn’t end well for ones who didnt)

When Pixar make the first 3d movie to take off. people where talking about how 3d animation would kill 2d.

11 years later after Disney switched to 3d we still got 2d movies and then people use 3d and 2d tricks make the show the dragon prince.

ai art is killing art

photoshop is killing photograph

3d amination are killing animation.

Computers are killing hand drawn art.

the power loom put all weavers out of work

None of the comparisons you make applies to generative AI. The gAI tools can generate imagery without human help at en mass. All the other tools still need human operatives, they just speed up the process. While the gAI tools can be used by operatives, in the end they do not need anybody at this point. They can be scripted to generate millions of images in minutes, with no human help.

Bear in mind that major Internet outlets have started using GPT like tech to generate content, which means they have been firing staff slowly. The same transformation is happening in the production art field. It is good to be positive and upbeat about the future of mankind, but the reality on the ground has been changing very quickly in favor of automation in art and liberal arts.

5 Likes

Have you tired chat GPT or stable diffusion. the people who let it run without oversite will fail very quickly.

unless you have a team(yes smaller then current teams now) that double checking the AIs back. you will get crap cookie cutter news posts with more misinformation then a fox news clip.

heck wow users have been messing with this AI bots alreadywith wow players tricking a bot into thinking glorbo is a thing im telling you chat GPT looks smart but dumb if given no human guidance.

also ai art can give me 1000 renders of cats but if I wanna render my cat with matching fur pattens and rooms context and pose i want. I’m going have to collect and render extra data plug into the control net followed by hour or two of picking into the data.

I’m also at disadvantage over non ai artist snice I suck at 2d I have less of an eye for it so while barely keep up with the 2d artist with the. when the artists give in and use it they would be so far ahead of me Ill be hopeless to catch up

The big is issue is artist and tech bros are screaming at each other while not learning each side of their story. but the big think Im worried about is the some of tech bro who wish make their ai art better with start studying and learning 2d art concepts that the AI lacks then will soon learn how to do things the non ai artist learned themselves. and since the non ai artists didnt start now learn will left in the dust.

and staff being fired a chunk of them will be sheepishy rehired when bean counters wonder why users feel about the lack luster artwork and news posts.

Imo AI images should show their sources all the time to be legal.
If 5 million images were used to build an image, all artists making that image possible should be credited.
That or pay a % to every creator that contributed to fuel their model, unless their model is based of copyright free stuff, of course.

That would be a one time pay i guess.
Iam no one who makes a living from art, but i think
if the AI takes a “make an landscape in the style of artist ‘x’”,
artist x should get payed for it every time or whatever the artists arranged with the AI owner.

Styles are not copyrightable.

I’ve been putting a little something together about consciousness and coming to a conclusion, as to A.I and mankind, but it was getting a little depressing, I will post it but I’m trying to think of an upside to end with, it’s not easy.

Anyhow, I thought I’d see what Emad Mostaque was talking about this week (see previous video above, We’ve lost control) and…((Big Sigh)) I wish I hadn’t.
In the first ten mins they (him and Tom Bilyeu) we’re talking about 3D models being created live, there will be no programmers and making a Hollywood style movie…FOR ONE.

They then go on to say…basically…a whole lot of things are going to be replaced. It’s a long vid but worth watching. You may only want to watch the first ten mins!! Do not watch if A.I is depressing you.

Interested in what you think. The A.I vs artists world is moving quickly…but that can also be said about many more jobs.

  • How AI Will DISRUPT The Entire World In 3 Years (Prepare Now While Others Panic)*
1 Like

Artists are also inspired by countless other artists. Wouldn’t it be fair and consequent if artists had to credit every other artist they were inspired by and compensate them in all their work?

2 Likes

This is how art works in its core. Artists do not exist in a vacuum. They get inspired. They create something, and they may inspire others to create.

Comparing AI to actual living and thinking artists is ridiculous.
AI is just a software, it can’t get inspired, it can’t make any logical conclusions, it only executes algorithm on patterns. It doesn’t deserve same rights as artists. Stop anthropomorphizing software.

Generative AI is built on exploitation by greedy companies baiting investors to shiny new toy. It isn’t even AI, because actual AI doesn’t exist at all.

Sorry, I did not read the whole of 800+ posts here. Could you explain what you mean with anthropomorphizing?

Extremely common arguments from AI bros is implying “AI” software was a human being and saying it’s “inspiring”, “referencing like artist”, “imagining new stuff” etc. They are trying to attribute extremely nuanced processes of human being into a machine learning software.

They get inspired, they learn from others. Why can’t I use a computer for something that is very similar to this?