MicrostockGroup Sponsors


Author Topic: Dall e 2 will make us all redundant?  (Read 18274 times)

0 Members and 1 Guest are viewing this topic.

« Reply #50 on: September 03, 2022, 04:37 »
+1
just started today - here are the results of my first session

mountain biker riding through an alpine meadow with mountains in the background



19th century naval battle 




I think your examples show something about how the AI works and the limitations it has.

It does not really understand what a mountain biker oder a meadow or a naval battle is. It creates images from the description with the help of the images it has available or was trained with. And for mountain bikers that will be mostly fotos and for 19. century naval battles it will be mainly fotos of oil paintings.

Therefor, when asked to create a picture of a 19. century battle, it looks like an oil painting. And as the AI does not understand what a naval battle is, it fails to show the battle aspect of the oil paintings it uses and we get just two ships sailing after each other.

Also the meadow looks really awful. It looks like an amalgam from various meadow pictures, but not like any meadow you would meet in reality. You see some colored blops, which are supposed to be flowers, but you cannot really recognize any specific flower or any blade of grass. The foreground is also conveniently out of focus, so that the AI does not have to create the details of the meadow. Not that the parts in focus look much better.

The mountain biker seems to just ride through the meadow, without any path or trail. Yet between the wheels and through the lower parts of the hind wheel, there appears a brown patch, indicating that the mountain biker was generated from pictures with a different background.
« Last Edit: September 03, 2022, 04:39 by Big Toe »


Brasilnut

  • Author Brutally Honest Guide to Microstock & Blog

« Reply #51 on: September 03, 2022, 06:44 »
0

« Reply #52 on: September 03, 2022, 06:51 »
0
It does not really understand what a mountain biker oder a meadow or a naval battle is.

To elaborate on that:

We as humans know that cannons firing or something else indicating fighting is very important for a picture of a naval battle.

It tells the spectator that they are watching a picture of a battle and not, say, a regatta or a chance meeting of ships.

The AI does not know that. For the AI the firing cannons are just minor details that can be lost. It does not get the essence of a picture of a naval battle.

« Reply #53 on: September 03, 2022, 13:09 »
0

I think your examples show something about how the AI works and the limitations it has.

It does not really understand what a mountain biker oder a meadow or a naval battle is. It creates images from the description with the help of the images it has available or was trained with. And for mountain bikers that will be mostly fotos and for 19. century naval battles it will be mainly fotos of oil paintings.

Therefor, when asked to create a picture of a 19. century battle, it looks like an oil painting. And as the AI does not understand what a naval battle is, it fails to show the battle aspect of the oil paintings it uses and we get just two ships sailing after each other.

Also the meadow looks really awful. It looks like an amalgam from various meadow pictures, but not like any meadow you would meet in reality. You see some colored blops, which are supposed to be flowers, but you cannot really recognize any specific flower or any blade of grass. The foreground is also conveniently out of focus, so that the AI does not have to create the details of the meadow. Not that the parts in focus look much better.

The mountain biker seems to just ride through the meadow, without any path or trail. Yet between the wheels and through the lower parts of the hind wheel, there appears a brown patch, indicating that the mountain biker was generated from pictures with a different background.


agreed - none of these are ready for prime time - current status of these AI recalls Samuel Johnson's quip
  "Sir, [deleted sexist comment] like a dog's walking on his hind legs. It is not done well; but you are surprised to find it done at all."

asking for roman legion attacking a city wall came close - showed a few roman soldiers attacking, and had the wall, but quality of the soldiers was low.  probably need to refine the ask

One thing is clear - better results when people not the main subject, which means models are safe for now

Fort now, there's more potential for changing the images into illustrations - here's a rough edit for long view of climbers near summit of an himalyan peak

« Reply #54 on: September 03, 2022, 13:10 »
+2
from today's NYTimes  An A.I.-Generated Picture Won an Art Prize. Artists Arent Happy.

https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html

Controversy over new art-making technologies is nothing new. Many painters recoiled at the invention of the camera, which they saw as a debasement of human artistry. (Charles Baudelaire, the 19th-century French poet and art critic, called photography arts most mortal enemy.) In the 20th century, digital editing tools and computer-assisted design programs were similarly dismissed by purists for requiring too little skill of their human collaborators.

« Reply #55 on: October 03, 2022, 06:11 »
+1
Dall-e 2 and other ai software are just what we need: more junk & crap images.



Brasilnut

  • Author Brutally Honest Guide to Microstock & Blog

« Reply #56 on: October 03, 2022, 08:38 »
+5
Last weekend I invested on an old green suitcase at a flea market and using Dalle2.

Finding it very useful to brainstorm various concepts for book covers.

« Last Edit: October 05, 2022, 06:20 by Brasilnut »

« Reply #57 on: October 03, 2022, 14:05 »
+2
Dall-e 2 and other ai software are just what we need: more junk & crap images.

Interesting your point of view... don't expect AI is like a jukebox where you push a button and that's it. It's not.
You need to figure out good prompts, train models or acquire good trained models (which depends of what are your needs) then... set up number of steps and among a big list of parameters that you can tweak.

Nevertheless i've been selling here and there Ai images/animations since 2018. For me AI is just a part of improving my performance of work. I just see AI for Image as a tool like Photoshop or a simple color pencil to help me create.



 


« Reply #58 on: October 03, 2022, 20:25 »
+4
Hey guys, i will sell all my photography and design gear, and i will  learn anothers skills. This AI monster probably will end with my extra income microstock. Sad but true! I will sell all my photography stuff, because wait so much would be so late! :-\

« Reply #59 on: October 04, 2022, 02:58 »
+2
Dall-e 2 and other ai software are just what we need: more junk & crap images.

Interesting your point of view... don't expect AI is like a jukebox where you push a button and that's it. It's not.
You need to figure out good prompts, train models or acquire good trained models (which depends of what are your needs) then... set up number of steps and among a big list of parameters that you can tweak.

Nevertheless i've been selling here and there Ai images/animations since 2018. For me AI is just a part of improving my performance of work. I just see AI for Image as a tool like Photoshop or a simple color pencil to help me create.

Why go all through that trouble when you can excel in photography or creating illustrations, learn and do great work on Photoshop and with other image editors yourself? If you have patience, skill and willingness to learn that is?

Besides, It is always a good idea to read the terms of service. You can't even be the copyright owner.

« Reply #60 on: October 04, 2022, 12:32 »
0
" Why go all through that trouble when you can excel in photography or creating illustrations, learn and do great work on Photoshop and with other image editors yourself? If you have patience, skill and willingness to learn that is? "

Simple answer would be: its faster.

Keep in mind that even Photoshop now uses AI too and It's not a "trouble" for me. it's just another tool to help me create artwork.
I start with analogue and moved to digital. i like both. But it's faster with digital and results depends of what i want to achieve in a creative sense. The same happens with 3d renders. In a way they are photography too although the process is not quite the same ....still.... renders calculate the light and reflect them into objects like real world. Nowadays i see a lot of people even using Unreal engine (game engine) to do photography artworks.


"Besides, It is always a good idea to read the terms of service. You can't even be the copyright owner."

It's funny you mention this... because in university there are studies that point that the Prompt is the new "digital signature" of the artist. You can copy/paste the prompt but the AI does not create equal images, like fingerprints (they are similar but not equal). This is arguable of course but not in the realm of authenticity but rather understanding that humans are now Human-machine people. AI is just another exponential of you and not a substitute of you, like a pen, a smartphone or your email account.

In terms of copyright owner this depends what models you use or how they are trained for your creative artwork. You will find free libraries of AI trained models on GitHub but you can create your own. the code is free. i suggest you to use google Colab for this.


So, in short: From Analogue to digital... from Photoshop to C4D/UE5 or AI tools/code similar to Dall-e  i use the tool that i find it will be better to create what i want to achieve.  ;)


« Reply #61 on: October 04, 2022, 13:45 »
+1
Last weekend I invested on an old green suitcase at a flea market and using Dalle2 very useful to brainstorm various concepts for book covers.

This is a GREAT way using AI! 👍👍👍

« Reply #62 on: October 04, 2022, 16:51 »
+4
...

Besides, It is always a good idea to read the terms of service. You can't even be the copyright owner.

did YOU read DALL E's TOS? it's been mentioned here before that they explicitly say the copyright belongs to the artist

« Reply #63 on: October 04, 2022, 18:18 »
+5
" Why go all through that trouble when you can excel in photography or creating illustrations, learn and do great work on Photoshop and with other image editors yourself? If you have patience, skill and willingness to learn that is? "

Simple answer would be: its faster.

Keep in mind that even Photoshop now uses AI too and It's not a "trouble" for me. it's just another tool to help me create artwork.
I start with analogue and moved to digital. i like both. But it's faster with digital and results depends of what i want to achieve in a creative sense. The same happens with 3d renders. In a way they are photography too although the process is not quite the same ....still.... renders calculate the light and reflect them into objects like real world. Nowadays i see a lot of people even using Unreal engine (game engine) to do photography artworks.


"Besides, It is always a good idea to read the terms of service. You can't even be the copyright owner."

It's funny you mention this... because in university there are studies that point that the Prompt is the new "digital signature" of the artist. You can copy/paste the prompt but the AI does not create equal images, like fingerprints (they are similar but not equal). This is arguable of course but not in the realm of authenticity but rather understanding that humans are now Human-machine people. AI is just another exponential of you and not a substitute of you, like a pen, a smartphone or your email account.

In terms of copyright owner this depends what models you use or how they are trained for your creative artwork. You will find free libraries of AI trained models on GitHub but you can create your own. the code is free. i suggest you to use google Colab for this.


So, in short: From Analogue to digital... from Photoshop to C4D/UE5 or AI tools/code similar to Dall-e  i use the tool that i find it will be better to create what i want to achieve.  ;)

I really wish I could look at this as just another tool in my kit, but the singularitarian in me fails to see how this can possibly benefit myself at the moment let alone humanity in the long term. Right now its still just another disruptive technology poised to benefit a select few at the expense of many. Nothing new under the sun there. But just another tool it is most definitely not. The speed with which it is encroaching on human expression and creativity is alarming to me. And the fact that (as predicted) it is developing so rapidly and is being so wholeheartedly embraced by so many without thoughtful consideration to its real impact is distressing. In the end though Im curious what it is you think that you will actually be achieving? Certainly nothing special IMO. And once the technology has matured and become so ubiquitous and pervasive that everyone will be using it for everything will your digital graffiti still be original, noteworthy, or even remotely interesting to anyone other than yourself? I highly doubt it. Nor will it be marketable. I believe that what you and others are forgetting is that the sole evolutionary imperative of AI is to ultimately make the you part of the equation irrelevant. You yourself mentioned the concept of human-machine people but never mentioned the fact that after the singularity the next purely logical evolution doesnt bode well for humanity. Remove art, expression, individuality and perhaps most importantly trust from the equation and you are hastening the process.

« Reply #64 on: October 05, 2022, 09:34 »
0
"Remove art, expression, individuality and perhaps most importantly trust from the equation and you are hastening the process."

I share some of your point of views specially what are the boundaries of Art.  Where exactly is the line or is there any? What is art? What defines an artist? questions that an artwork should reflect.

The answers to these questions were shaken in 1917, when Frenchman Marcel Duchamp brought a urinal  signed R. Mutt to an exhibition by the New York Association of Independent Artists. The condition to exhibit was simple: pay 6 dollars for registration. Duchamp would called a ready-made (already ready). Banal objects turned into art. What makes it unique was the signature.

Others follow this ready made art path like Andy Warhol with Pop Art. I don't think both artist remove art, expression, individuality. Actually IMO they are the added value.

Exactly what you would like to achieve with your work. Art or just being creative i trust in your feelings and judgment.








« Reply #65 on: October 05, 2022, 13:54 »
0
..
...You yourself mentioned the concept of human-machine people but never mentioned the fact that after the singularity the next purely logical evolution doesnt bode well for humanity. Remove art, expression, individuality and perhaps most importantly trust from the equation and you are hastening the process.

the next steps will see AI for stock buyers, then they'll replace graphic designers.  AI will read & post to social media and decide what their humans (a la 'Mr Peabody's boy sherman') 'want' to buy

i mentioned earlier a thoughtful take:
https://www.cold-takes.com/ai-could-defeat-all-of-us-combined/

« Reply #66 on: October 06, 2022, 02:39 »
+4
..
...You yourself mentioned the concept of human-machine people but never mentioned the fact that after the singularity the next purely logical evolution doesnt bode well for humanity. Remove art, expression, individuality and perhaps most importantly trust from the equation and you are hastening the process.

the next steps will see AI for stock buyers, then they'll replace graphic designers.  AI will read & post to social media and decide what their humans (a la 'Mr Peabody's boy sherman') 'want' to buy

i mentioned earlier a thoughtful take:
https://www.cold-takes.com/ai-could-defeat-all-of-us-combined/

What I find interesting is what will people do to earn money once AI is encouraged to replace all the jobs. Yes I know new jobs will recreated but nowhere near as many. AI is writing books, news articles and the like and that's before things like automated cars etc become the norm. Thousands, millions of jobs gone. If a large % of the population is no longer earning money (or as much money) who is going to buy and use the services. Given the amount of price cutting in all areas in order to get an edge, a reduction in sales is the last thing that is needed.

I know this is somewhat "me me me" but I'm glad I'm a lot closer to retirement than most around here. The future will certainly be a lot different and I feel there is a rush from some quarters to embrace the end of human involvement in the creative process. It's not just a tool, it's creating a system that removes the person from the creative process. The end result will be a client and the software. It won't take many people to manage the software for a world wide audience and the client once they get used to it won't need someone to type in the words for them.

By the time they realise that part of the creative process was the actual thinking of the ideas, millions of people will have left the industry and they'll be left with a piece of software that finds it hard to think outside the box... because it doesn't think. It scraps data (visual and words) and try to learn from it. What happens when there is no one left to create fresh data for it to learn from?!? Will it just end up learning from images it's created itself, including the errors it makes as it can't think or understand what the real item should look like... other than from the scrapped data. Will it be able to come up with new trends?!? Invent different styles. So far, the images it produces seem very similar in style. So much so I tend to be able to spot them very quickly online. How long before people get bored? Hopefully I'll be sitting with my feet up in front of the fire enjoying a single malt (retired) by that time.
« Last Edit: October 06, 2022, 02:46 by HalfFull »

tupungato

  • Europe
« Reply #67 on: October 06, 2022, 03:46 »
+3
The problem with current AI is the questionable legal status of training data. It just crawls and harvests the internet, right? So, incorporating copyrighted elements, parts of someone else's artwork is inevitable.
I'm already battle hardened in copyright legal battles (as a creator whose work was infringed upon). A lot of my images are default results in Google's first pages for many keywords. Bring it on!

« Reply #68 on: October 06, 2022, 04:07 »
+2
incorporating copyrighted elements, parts of someone else's artwork is inevitable

AI doesn't incorporate anything.
AI learn what is and how to recreate any object (or human faces, animals... everything)

Of course there are legal problems because images used to train are copyrighted; but there is nothing that will be "incorporated" in new images

It's quite new scheme, and it cannot be managed with "classic" discussion, it's completely new issue to solve.

« Reply #69 on: October 06, 2022, 06:04 »
+3
incorporating copyrighted elements, parts of someone else's artwork is inevitable

AI doesn't incorporate anything.
AI learn what is and how to recreate any object (or human faces, animals... everything)

Of course there are legal problems because images used to train are copyrighted; but there is nothing that will be "incorporated" in new images

It's quite new scheme, and it cannot be managed with "classic" discussion, it's completely new issue to solve.

Interestingly, it obviously copies quite a bit as they were also including watermarks with the images they produce. Of course, the programmers will write a bit of code to remove them in the future but it's obvious it's basing images on real content.

« Reply #70 on: October 06, 2022, 06:24 »
+2



Interestingly, it obviously copies quite a bit as they were also including watermarks with the images they produce.


Might risk sounding like a broken record, but: The AIs sometimes generated images that have something resembling microstock agency watermarks, because they have been trained with so many watermarked (unlicensed!) images that they wrongly learned that the watermark was part of whatever it was supposed to generate. When an AI generates a watermark, it "thinks" it belongs in the picture like a suit to a businessman or the sun to a picture of a sunny sky. It's an issue of wrong learning, not an issue of copying. It recreates the watermark, just like it re-creates the sun or a suit. It cannot understand that the watermark is not part of whatever it is supposed to depict. If an AI was capable of thinking/realizing that whatever it is creating in images was actually something that exists in the offline world, then it would think that people walk around with floating watermarks in front of them.

I start to think that many people do not really understand what an AI is. Artificial intelligence. It's not a computer programm that copy & pastes stuff. It is a program that has learning abilities. It gets input and it learns from it. Give it the wrong input and it will learn to create wrong results.
« Last Edit: October 06, 2022, 06:27 by Her Ugliness »

« Reply #71 on: October 06, 2022, 06:41 »
+3



Interestingly, it obviously copies quite a bit as they were also including watermarks with the images they produce.


Might risk sounding like a broken record, but: The AIs sometimes generated images that have something resembling microstock agency watermarks, because they have been trained with so many watermarked (unlicensed!) images that they wrongly learned that the watermark was part of whatever it was supposed to generate. When an AI generates a watermark, it "thinks" it belongs in the picture like a suit to a businessman or the sun to a picture of a sunny sky. It's an issue of wrong learning, not an issue of copying. It recreates the watermark, just like it re-creates the sun or a suit. It cannot understand that the watermark is not part of whatever it is supposed to depict. If an AI was capable of thinking/realizing that whatever it is creating in images was actually something that exists in the offline world, then it would think that people walk around with floating watermarks in front of them.

I start to think that many people do not really understand what an AI is. Artificial intelligence. It's not a computer programm that copy & pastes stuff. It is a program that has learning abilities. It gets input and it learns from it. Give it the wrong input and it will learn to create wrong results.

As someone who used to use machine learning for building credit scorecards I'm well aware of what is and how it learns but the point here is it's learned from using our copyright material without consent and with no real way of guaranteeing it doesn't create an image with more than a passing resemblance to others work. But hey, if you and others don't see a problem with this it's your future. The only way we could build scorecards via machine learning was to provide it the raw material from existing behaviour along lists of criteria we would tweak in order for it to produce what we wanted.

I understand a lot of people who aren't illustrators will love it as they can now supply material but this will only be short term as clients will cut out this middleman in the end.
« Last Edit: October 06, 2022, 06:44 by HalfFull »

« Reply #72 on: October 06, 2022, 08:19 »
0
Interesting analysis from Fabio Nodari here:

https://www.fabionodariphoto.com/en/how-sell-ai-generated-images-art/

Flawed image because of the reflections of things that aren't there. Anyone looking for a few seconds would say, it looks fake and wrong.

« Reply #73 on: October 06, 2022, 09:27 »
+2
It is a program that has learning abilities. It gets input and it learns from it.

Like kids learning in a school with stolen books, stolen lab materials, stolen utilities, etc

Once the kids have learned the skill, they move on and create new original stuff, as instructed by their employer.
« Last Edit: October 06, 2022, 10:49 by Zero Talent »

« Reply #74 on: October 06, 2022, 10:57 »
+6
If i want to learn to drive a car, could i steal a car for two days and gave it back when i learned? If i want to fly a plane, could i steal for one month and gave it back when i learned?


 

Related Topics

  Subject / Started by Replies Last post
11 Replies
4901 Views
Last post August 08, 2013, 02:12
by anjaliaroha
3 Replies
2666 Views
Last post May 28, 2014, 16:50
by gnirtS
19 Replies
5340 Views
Last post May 18, 2022, 05:21
by thx9000
14 Replies
3243 Views
Last post October 20, 2022, 04:46
by MadMax
13 Replies
3417 Views
Last post April 06, 2023, 12:26
by Uncle Pete

Sponsors

Mega Bundle of 5,900+ Professional Lightroom Presets

Microstock Poll Results

Sponsors