AI Generated Art

Tech

September 13, 2022

LEAVE A COMMENT

If you surf the interwebs as much as me, your eyeballs have been assaulted by the volume of Ai generated art; be it the meme worthy craiyon image of Bruce Willis eating yoghurt or perhaps you heard that a midjourney creation just fooled some judges and took out a digital art competition .  Cosmopolitan magazine decided that graphic designers were redundant and enlisted Karen X Cheng to create a recent cover.  There is a backlash, as the freelance journalist Charlie Warzel discovered when he included two pieces of Ai art in an article for the Atlantic.

The volume of commentary and fear mongering is becoming overwhelming, with the overarching theme of Ai will take my job. The technofuturists try to temper fear with think pieces such as please don’t fear the Ai.

Artist David O’Reily says:

Whenever a skill becomes automated, art moves into new terrains of impossibility. The camera did to painting what guns did to spears. There is no going back.

Everyday we’re seeing crazier and more ambitious imagery as the arms race of AI models race to be the one to rule them all. For a breakdown of the pros and cons of each check out this analysis from PetaPixel. or the 10 best AI art generators.

Ai generated art - the Survey

I surveyed dozens of motion designers and 3D artists who are mostly based in Australia.

I found that most artists were using Midjourney, Dall-e2 or Stable diffusion with over 2/3rds dabbling in 2 or more different platfoms.  A quarter of people were already using it as part of their creative process, with another ⅓ open to using it in the future.

Half of those surveyed were primarily using it for Idea generation, a 3rd styleframing but mostly for mindless noodling.

Artists were primarily excited about the technology without worry although a large number were on the fence that AI generated art was going to effect their job.  There were a number of people who were passionately opposed and that AI was a real concern.

“From what I observe, one thing that has contributed to the longevity of my career is the ability to work with clients. To provide a service, makes them feel heard, and makes them feel they're satisfied with the solutions you provide. So as long as you can be that person, you'll never be out of a job. In that sense, both you and the client can and probably will use AI as a tool to get to the final visual solution that they need/want, but in the end, humans will (I believe) always prefer working with other humans.”

- Lilian Darmono

Artist turned designer Nando Costa is already using AI in his workflow experimenting with branding for coke and google, he muses in this article on Linkedin:

“What if machine learning could enable design studios to train an algorithm to act as an extension of their teams, and bring in AI as a tool to effectively shrink the early ideation process dramatically. During this stage, the firming up of a director's perspective on details such as composition, lighting, and other nuances before a photoshoot, or CG production starts could shrink from weeks or days, to hours.”

Prompting

Prompting, Prompt engineering or Prompt AI whispering, is the ‘art’ of formulating the bespoke series of words and phrases to yield desirable or useful results.  

If stringing together some descriptive words is too difficult there are some free guides out there including Phraser or Lexica which show examples of outputs including their text prompts.  The tool Prompter (for midjourney) is a highly detailed multiple choice form allowing the user to cook up text recipes with specific granularity.  There’s a plethora of sites to help guide novices what styles can be copied too such as Generated.

Of course some companies are trying to make a buck, such as PromptBase, promptoMANIA to take guesswork out of the secret sauce needed to create desirable imagery.

Ethics

Big data using the trove of internet content as raw material does bring up ethical questions, particularly those of living artists work is scraped in order to create new bespoke images in their style. This is a tricky area as a style can’t be copyrighted.  

Many contributors were concerned as to the ethics of the raw material used to create Ai artwork. Fantasy artist Greg Rutkowski ‘s name has been used as a prompt around 93,000 times on the platform Stable Diffusion, and understandably he’s extremely concerned.

- Illustrator Jono Yuen:

 ” The idea of using AI to clone a style and leveraging off the art that someone has developed over many years doesn’t sit well with me.”

The torrent of Ai generated art is being rejected by many art communities, at the time of writing, Newgrounds, Inkblot Art and Fur Affinity are all curbing the use of Ai art. Getty images, iStock and Unsplash are no longer accepting Ai images due to the origin of the source material, images are trained on.

- Illustrator, Glendon Mellow tweets:

“I am not okay with my artwork being included in word-prompt AI image databases without my consent and permission. It’s not ethical. These are powerful tools being built recklessly. “

As to who owns the copyright of Ai generated imagery is also murky. As for Dall-e2, you can use the images commercially but the company Open AI owns them. As for midjourney, for paid users, you own the copyright. Nevertheless Although under US law, images can only be copyrighted if created by a human. Legalities will no doubt need stricter definition in the months and years to come.

The Future

The train has left the station.  That train happens to be developed from millions of trains and millions of stations and there is no stopping them.  Ai has been implemented under our noses for years, it seems it's just now that we’ve started to take notice.

Half of the people I surveyed were already harnessing Ai in some capacity and it’s growing;.  integrating it in their workflows, speeding up the ideation process, selling NFTs or images on stock websites.

From a technology perspective I’m excited.  These are such powerful tools which are being improved on and adapted everyday; photographers using Inpainting to make blurry photos sharp , architects harnessing Ai to Imagining new buildings , designers developing  new sneakers and seamless texture creation.  I’m looking forward to photoshop integration, (adobe is already teasing some AI neural filters) but in the meantime, there’s a plugin harnessing Stable Diffusion.

From an artistic point of view I think its complicated.  Ai art derived from a string of words while surprising and magical can feels pretty soulless and like lots of people is wearing thin, like the fad of mocap dancers of a few years ago.

I’m also concerned that it will devalue the craft of the millions of creatives who have been making a living as creators. The illustrators, concept artists and stock photographers will find the next few years transition challenging and will need to adapt when some clients can make similar imagery themselves.  

The future looks exciting and kinda frightening at the same time.

- Tom

“Coming up with visual ideas to me was always the fun part of the process. Why do we need AI for this part of the process? I feel as though AI image generation is the product of a problem that didn't really exist.”

- Director, Andrew Daffy

“The sooner we treat AI as a fellow artist, contributor, the sooner we move to a healthy future.”

- Creative Director, Jess Wheeler

“Where this will go, we don't know, but at this stage a lot of the panic is reminiscent of every other technological advance in the creative field.”

- Director - Jeremy Mansford

“Concerned a little about how it might steal away value from jobs or projects. Will clients expect more for less for certain processes now that AI could do it?”

- Project Manager, Luke Gee:

Does it remove the craft from art or create a new type of craft?”

- Director - Karen x Cheng

“I think the natural reaction is to fear that AI will replace human artists. Certainly, that thought crossed my mind, especially in the beginning. But the more I use DALL-E, the less I see this as a replacement for humans, and the more I see it as tool for humans to use — an instrument to play.”

Words by Simon Bronson
Illustrations by Annika Fleur

Continue the conversation on the Node slack #AI channel

Request an invite here

Further reading & links

OpenArt - ai gallery

AI Images: Last Week Tonight with John Oliver (HBO)

Dall-e 2 beta

The Dall-e2 Prompt book

Midjourney Prompt help guide

I replaced all our blog thumbnails using DALL·E 2 for $45: here’s what I learned

Replicate - free AI models to try

Opinion: OpenAI’s DALL-E 2 is the big tech equivalent of ‘soylent green’


MidJourney Neural Algorithm Sharpens Style Of AI Art 

Will This New Invention be the Death of Photography? 

Designing Websites with DALL-E 2 

Getting started with Disco Diffusion 

Disco Diffusion 70+ Artist Studies 

Nando Costa on co-creating with AI 

Marquis Brownlee's designer goes head to head with AI 

Copyright-infringement-in-artificial-intelligence-art

Heading

LEAVE A COMMENT

September 13, 2022

If you surf the interwebs as much as me, your eyeballs have been assaulted by the volume of Ai generated art; be it the meme worthy craiyon image of Bruce Willis eating yoghurt or perhaps you heard that a midjourney creation just fooled some judges and took out a digital art competition .  Cosmopolitan magazine decided that graphic designers were redundant and enlisted Karen X Cheng to create a recent cover.  There is a backlash, as the freelance journalist Charlie Warzel discovered when he included two pieces of Ai art in an article for the Atlantic.

The volume of commentary and fear mongering is becoming overwhelming, with the overarching theme of Ai will take my job. The technofuturists try to temper fear with think pieces such as please don’t fear the Ai.

Artist David O’Reily says:

Whenever a skill becomes automated, art moves into new terrains of impossibility. The camera did to painting what guns did to spears. There is no going back.

Everyday we’re seeing crazier and more ambitious imagery as the arms race of AI models race to be the one to rule them all. For a breakdown of the pros and cons of each check out this analysis from PetaPixel. or the 10 best AI art generators.

Ai generated art - the Survey

I surveyed dozens of motion designers and 3D artists who are mostly based in Australia.

I found that most artists were using Midjourney, Dall-e2 or Stable diffusion with over 2/3rds dabbling in 2 or more different platfoms.  A quarter of people were already using it as part of their creative process, with another ⅓ open to using it in the future.

Half of those surveyed were primarily using it for Idea generation, a 3rd styleframing but mostly for mindless noodling.

Artists were primarily excited about the technology without worry although a large number were on the fence that AI generated art was going to effect their job.  There were a number of people who were passionately opposed and that AI was a real concern.

“From what I observe, one thing that has contributed to the longevity of my career is the ability to work with clients. To provide a service, makes them feel heard, and makes them feel they're satisfied with the solutions you provide. So as long as you can be that person, you'll never be out of a job. In that sense, both you and the client can and probably will use AI as a tool to get to the final visual solution that they need/want, but in the end, humans will (I believe) always prefer working with other humans.”

- Lilian Darmono

Artist turned designer Nando Costa is already using AI in his workflow experimenting with branding for coke and google, he muses in this article on Linkedin:

“What if machine learning could enable design studios to train an algorithm to act as an extension of their teams, and bring in AI as a tool to effectively shrink the early ideation process dramatically. During this stage, the firming up of a director's perspective on details such as composition, lighting, and other nuances before a photoshoot, or CG production starts could shrink from weeks or days, to hours.”

Prompting

Prompting, Prompt engineering or Prompt AI whispering, is the ‘art’ of formulating the bespoke series of words and phrases to yield desirable or useful results.  

If stringing together some descriptive words is too difficult there are some free guides out there including Phraser or Lexica which show examples of outputs including their text prompts.  The tool Prompter (for midjourney) is a highly detailed multiple choice form allowing the user to cook up text recipes with specific granularity.  There’s a plethora of sites to help guide novices what styles can be copied too such as Generated.

Of course some companies are trying to make a buck, such as PromptBase, promptoMANIA to take guesswork out of the secret sauce needed to create desirable imagery.

Ethics

Big data using the trove of internet content as raw material does bring up ethical questions, particularly those of living artists work is scraped in order to create new bespoke images in their style. This is a tricky area as a style can’t be copyrighted.  

Many contributors were concerned as to the ethics of the raw material used to create Ai artwork. Fantasy artist Greg Rutkowski ‘s name has been used as a prompt around 93,000 times on the platform Stable Diffusion, and understandably he’s extremely concerned.

- Illustrator Jono Yuen:

 ” The idea of using AI to clone a style and leveraging off the art that someone has developed over many years doesn’t sit well with me.”

The torrent of Ai generated art is being rejected by many art communities, at the time of writing, Newgrounds, Inkblot Art and Fur Affinity are all curbing the use of Ai art. Getty images, iStock and Unsplash are no longer accepting Ai images due to the origin of the source material, images are trained on.

- Illustrator, Glendon Mellow tweets:

“I am not okay with my artwork being included in word-prompt AI image databases without my consent and permission. It’s not ethical. These are powerful tools being built recklessly. “

As to who owns the copyright of Ai generated imagery is also murky. As for Dall-e2, you can use the images commercially but the company Open AI owns them. As for midjourney, for paid users, you own the copyright. Nevertheless Although under US law, images can only be copyrighted if created by a human. Legalities will no doubt need stricter definition in the months and years to come.

The Future

The train has left the station.  That train happens to be developed from millions of trains and millions of stations and there is no stopping them.  Ai has been implemented under our noses for years, it seems it's just now that we’ve started to take notice.

Half of the people I surveyed were already harnessing Ai in some capacity and it’s growing;.  integrating it in their workflows, speeding up the ideation process, selling NFTs or images on stock websites.

From a technology perspective I’m excited.  These are such powerful tools which are being improved on and adapted everyday; photographers using Inpainting to make blurry photos sharp , architects harnessing Ai to Imagining new buildings , designers developing  new sneakers and seamless texture creation.  I’m looking forward to photoshop integration, (adobe is already teasing some AI neural filters) but in the meantime, there’s a plugin harnessing Stable Diffusion.

From an artistic point of view I think its complicated.  Ai art derived from a string of words while surprising and magical can feels pretty soulless and like lots of people is wearing thin, like the fad of mocap dancers of a few years ago.

I’m also concerned that it will devalue the craft of the millions of creatives who have been making a living as creators. The illustrators, concept artists and stock photographers will find the next few years transition challenging and will need to adapt when some clients can make similar imagery themselves.  

The future looks exciting and kinda frightening at the same time.

- Tom

“Coming up with visual ideas to me was always the fun part of the process. Why do we need AI for this part of the process? I feel as though AI image generation is the product of a problem that didn't really exist.”

- Director, Andrew Daffy

“The sooner we treat AI as a fellow artist, contributor, the sooner we move to a healthy future.”

- Creative Director, Jess Wheeler

“Where this will go, we don't know, but at this stage a lot of the panic is reminiscent of every other technological advance in the creative field.”

- Director - Jeremy Mansford

“Concerned a little about how it might steal away value from jobs or projects. Will clients expect more for less for certain processes now that AI could do it?”

- Project Manager, Luke Gee:

Does it remove the craft from art or create a new type of craft?”

- Director - Karen x Cheng

“I think the natural reaction is to fear that AI will replace human artists. Certainly, that thought crossed my mind, especially in the beginning. But the more I use DALL-E, the less I see this as a replacement for humans, and the more I see it as tool for humans to use — an instrument to play.”

Words by Simon Bronson
Illustrations by Annika Fleur

Continue the conversation on the Node slack #AI channel

Request an invite here

Further reading & links

OpenArt - ai gallery

AI Images: Last Week Tonight with John Oliver (HBO)

Dall-e 2 beta

The Dall-e2 Prompt book

Midjourney Prompt help guide

I replaced all our blog thumbnails using DALL·E 2 for $45: here’s what I learned

Replicate - free AI models to try

Opinion: OpenAI’s DALL-E 2 is the big tech equivalent of ‘soylent green’


MidJourney Neural Algorithm Sharpens Style Of AI Art 

Will This New Invention be the Death of Photography? 

Designing Websites with DALL-E 2 

Getting started with Disco Diffusion 

Disco Diffusion 70+ Artist Studies 

Nando Costa on co-creating with AI 

Marquis Brownlee's designer goes head to head with AI 

Copyright-infringement-in-artificial-intelligence-art

Tech

September 13, 2022

LEAVE A COMMENT

If you surf the interwebs as much as me, your eyeballs have been assaulted by the volume of Ai generated art; be it the meme worthy craiyon image of Bruce Willis eating yoghurt or perhaps you heard that a midjourney creation just fooled some judges and took out a digital art competition .  Cosmopolitan magazine decided that graphic designers were redundant and enlisted Karen X Cheng to create a recent cover.  There is a backlash, as the freelance journalist Charlie Warzel discovered when he included two pieces of Ai art in an article for the Atlantic.

The volume of commentary and fear mongering is becoming overwhelming, with the overarching theme of Ai will take my job. The technofuturists try to temper fear with think pieces such as please don’t fear the Ai.

Artist David O’Reily says:

Whenever a skill becomes automated, art moves into new terrains of impossibility. The camera did to painting what guns did to spears. There is no going back.

Everyday we’re seeing crazier and more ambitious imagery as the arms race of AI models race to be the one to rule them all. For a breakdown of the pros and cons of each check out this analysis from PetaPixel. or the 10 best AI art generators.

Ai generated art - the Survey

I surveyed dozens of motion designers and 3D artists who are mostly based in Australia.

I found that most artists were using Midjourney, Dall-e2 or Stable diffusion with over 2/3rds dabbling in 2 or more different platfoms.  A quarter of people were already using it as part of their creative process, with another ⅓ open to using it in the future.

Half of those surveyed were primarily using it for Idea generation, a 3rd styleframing but mostly for mindless noodling.

Artists were primarily excited about the technology without worry although a large number were on the fence that AI generated art was going to effect their job.  There were a number of people who were passionately opposed and that AI was a real concern.

“From what I observe, one thing that has contributed to the longevity of my career is the ability to work with clients. To provide a service, makes them feel heard, and makes them feel they're satisfied with the solutions you provide. So as long as you can be that person, you'll never be out of a job. In that sense, both you and the client can and probably will use AI as a tool to get to the final visual solution that they need/want, but in the end, humans will (I believe) always prefer working with other humans.”

- Lilian Darmono

Artist turned designer Nando Costa is already using AI in his workflow experimenting with branding for coke and google, he muses in this article on Linkedin:

“What if machine learning could enable design studios to train an algorithm to act as an extension of their teams, and bring in AI as a tool to effectively shrink the early ideation process dramatically. During this stage, the firming up of a director's perspective on details such as composition, lighting, and other nuances before a photoshoot, or CG production starts could shrink from weeks or days, to hours.”

Prompting

Prompting, Prompt engineering or Prompt AI whispering, is the ‘art’ of formulating the bespoke series of words and phrases to yield desirable or useful results.  

If stringing together some descriptive words is too difficult there are some free guides out there including Phraser or Lexica which show examples of outputs including their text prompts.  The tool Prompter (for midjourney) is a highly detailed multiple choice form allowing the user to cook up text recipes with specific granularity.  There’s a plethora of sites to help guide novices what styles can be copied too such as Generated.

Of course some companies are trying to make a buck, such as PromptBase, promptoMANIA to take guesswork out of the secret sauce needed to create desirable imagery.

Ethics

Big data using the trove of internet content as raw material does bring up ethical questions, particularly those of living artists work is scraped in order to create new bespoke images in their style. This is a tricky area as a style can’t be copyrighted.  

Many contributors were concerned as to the ethics of the raw material used to create Ai artwork. Fantasy artist Greg Rutkowski ‘s name has been used as a prompt around 93,000 times on the platform Stable Diffusion, and understandably he’s extremely concerned.

- Illustrator Jono Yuen:

 ” The idea of using AI to clone a style and leveraging off the art that someone has developed over many years doesn’t sit well with me.”

The torrent of Ai generated art is being rejected by many art communities, at the time of writing, Newgrounds, Inkblot Art and Fur Affinity are all curbing the use of Ai art. Getty images, iStock and Unsplash are no longer accepting Ai images due to the origin of the source material, images are trained on.

- Illustrator, Glendon Mellow tweets:

“I am not okay with my artwork being included in word-prompt AI image databases without my consent and permission. It’s not ethical. These are powerful tools being built recklessly. “

As to who owns the copyright of Ai generated imagery is also murky. As for Dall-e2, you can use the images commercially but the company Open AI owns them. As for midjourney, for paid users, you own the copyright. Nevertheless Although under US law, images can only be copyrighted if created by a human. Legalities will no doubt need stricter definition in the months and years to come.

The Future

The train has left the station.  That train happens to be developed from millions of trains and millions of stations and there is no stopping them.  Ai has been implemented under our noses for years, it seems it's just now that we’ve started to take notice.

Half of the people I surveyed were already harnessing Ai in some capacity and it’s growing;.  integrating it in their workflows, speeding up the ideation process, selling NFTs or images on stock websites.

From a technology perspective I’m excited.  These are such powerful tools which are being improved on and adapted everyday; photographers using Inpainting to make blurry photos sharp , architects harnessing Ai to Imagining new buildings , designers developing  new sneakers and seamless texture creation.  I’m looking forward to photoshop integration, (adobe is already teasing some AI neural filters) but in the meantime, there’s a plugin harnessing Stable Diffusion.

From an artistic point of view I think its complicated.  Ai art derived from a string of words while surprising and magical can feels pretty soulless and like lots of people is wearing thin, like the fad of mocap dancers of a few years ago.

I’m also concerned that it will devalue the craft of the millions of creatives who have been making a living as creators. The illustrators, concept artists and stock photographers will find the next few years transition challenging and will need to adapt when some clients can make similar imagery themselves.  

The future looks exciting and kinda frightening at the same time.

- Tom

“Coming up with visual ideas to me was always the fun part of the process. Why do we need AI for this part of the process? I feel as though AI image generation is the product of a problem that didn't really exist.”

- Director, Andrew Daffy

“The sooner we treat AI as a fellow artist, contributor, the sooner we move to a healthy future.”

- Creative Director, Jess Wheeler

“Where this will go, we don't know, but at this stage a lot of the panic is reminiscent of every other technological advance in the creative field.”

- Director - Jeremy Mansford

“Concerned a little about how it might steal away value from jobs or projects. Will clients expect more for less for certain processes now that AI could do it?”

- Project Manager, Luke Gee:

Does it remove the craft from art or create a new type of craft?”

- Director - Karen x Cheng

“I think the natural reaction is to fear that AI will replace human artists. Certainly, that thought crossed my mind, especially in the beginning. But the more I use DALL-E, the less I see this as a replacement for humans, and the more I see it as tool for humans to use — an instrument to play.”

Words by Simon Bronson
Illustrations by Annika Fleur

Continue the conversation on the Node slack #AI channel

Request an invite here

Further reading & links

OpenArt - ai gallery

AI Images: Last Week Tonight with John Oliver (HBO)

Dall-e 2 beta

The Dall-e2 Prompt book

Midjourney Prompt help guide

I replaced all our blog thumbnails using DALL·E 2 for $45: here’s what I learned

Replicate - free AI models to try

Opinion: OpenAI’s DALL-E 2 is the big tech equivalent of ‘soylent green’


MidJourney Neural Algorithm Sharpens Style Of AI Art 

Will This New Invention be the Death of Photography? 

Designing Websites with DALL-E 2 

Getting started with Disco Diffusion 

Disco Diffusion 70+ Artist Studies 

Nando Costa on co-creating with AI 

Marquis Brownlee's designer goes head to head with AI 

Copyright-infringement-in-artificial-intelligence-art

```