Podcast: Download (Duration: 1:13:05 — 58.7MB)
Subscribe: Google Podcasts | Spotify | Stitcher | TuneIn | RSS | More
How can you use AI tools to ethically and responsibly create in whatever sphere you love? What are some of the tools and why are creative direction, voice, and taste, so important? I discuss these issues and more in a solo introduction and an interview with Oliver Altair.
In the first 28 mins of the podcast, I give an extended introduction about the various legal cases around AI and copyright, my thoughts on the best way to approach it for your creative work, and how to use AI tools ethically and responsibly. I've included the transcript below with lots of links and further resources, and you can find more at TheCreativePenn.com/future.
If you'd like to learn more, you can get 50% off my ebook and audiobook on Artificial Intelligence, Blockchain, and Virtual Worlds if you buy direct, and you can get 50% off my course, The AI-Assisted Author. Just use discount code: FEB23 at checkout for either.
Today's show is sponsored by my wonderful patrons who fund my brain so I have time to think about and discuss these futurist topics impacting authors. If you support the show, you also get the extra monthly patron-only Q&A audio. You can support the show at www.patreon.com/thecreativepenn
Oliver Altair is a dark fantasy author, a digital artist and creator of the Ravensfield Collection.
You can listen above or on your favorite podcast app or read the notes and links below. Here are the highlights and the full transcript is below. The interview starts at 28:33 mins.
- The creation of the Ravensfield Collection using AI art
- How AI art can be used as a marketing tool for authors
- Available tools for an AI-augmented creative
- How to write AI prompts to achieve the best results
- Concerns about intellectual property when training models
- What are the different kinds of Creative Commons licenses?
- The importance of creative confidence, voice, and taste, for making art with AI tools
- Generative writing tools as a brainstorming co-pilot
You can find Oliver Altair at OliverAltair.com. You can view the Ravensfield Collection at Ravensfield.art.
Header image by Joanna Penn on Midjourney.
Transcript of the solo introduction on generative AI with Joanna Penn
Before we get into the interview, I just want to add some over-arching comments as I’m getting a lot of emails about generative AI for art and obviously for words.
It feels like a lot of people are going through now what I did a few years back when I first heard about all this, so I understand how it feels to be kind of scared, kind of excited, kind of unsure as to what we can do with all this, and of course, you have to spend some time figuring out your position on it all and that’s not helped by the hyperbole, misunderstanding, and rage fuelled by press and social media.
There is a real sense of fear and confusion in the media and amongst different kinds of creators.
But instead of jumping into the outrage, take a deep breath and do some research for yourself from different points of view — and of course, I am only one point of view!
As I have talked about many times over the last few years, these generative AI systems are tools, and new technology always enables new forms of creative expression and helps creators of all kinds achieve new things.
Photography didn’t kill painting or drawing, online gaming didn’t kill tabletop games, ebooks didn’t kill beautiful print editions, synthesised music didn’t kill the live concert. In fact, all these things made the personal touch even more important.
Yes, there are many issues with generative AI, I know it’s not all rainbows and unicorns! Tools can also be weapons, to borrow the title of Brad Smith’s book on the promise and peril of the digital age — which was more about the disruption of the internet which has a similar dichotomy — but we can use these tools responsibly and ethically and encourage others to do the same.
Let’s first take a step back
These tools are not new, they have just been noticed by people outside the tech industry since ChatGPT was launched in late November 2023.
I started commenting about generative AI on the show back in 2016, and have done many interviews on it since then — you can find all the backlist episodes at TheCreativePenn.com/future
In July 2019, I did a solo episode on 9 Ways That Artificial Intelligence will Disrupt Authors and the Publishing Industry in the next decade, episode 437 if you want to listen.
My very first point was “Non-fiction books, blog posts, and news articles will be written by AI,” which is certainly happening with GPT3, ChatGPT, tools like Jasper, and more.
I also said that “Copyright law will be challenged as books are used to train AIs which then produce work in the voice of established authors.”
The first legal cases around copyright law have now been brought by artists as their art is being used to train models that can be used to produce work in the voice of established artists — it’s essentially the same thing as I wrote about.
There are a few legal cases about generative AI right now
As ever, I am not a lawyer/attorney, just an author, audiobook narrator and podcaster. I have an interview coming in the next few months with an IP lawyer on all this so we’ll circle back to it as the year progresses.
Three artists have filed a class action lawsuit against Midjourney, Stable Diffusion and DreamUp for scraping their images and using their intellectual property without consent.
I recommend listening to the excellent Hard Fork podcast episode from 20 Jan, 2023 which has a segment with artist Sarah Andersen on the lawsuit. She says they are not looking to shut down the technology, but they want consent so artists can opt their images in for training, rather than opt-out as default. They also want credit if their work is used, and compensation for any harm or infringement, and also for licensing data for training.
Getty Images are also suing Stable Diffusion for unlawfully scraping images from its site. In an article on The Verge, Getty Images CEO, Craig Peters “compares the current legal landscape in the generative AI scene to the early days of digital music, where companies like Napster offered popular but illegal services before new deals were struck with license holders like music labels.
Peters said Getty was not interested in financial damages or stopping the development of AI art tools, but in creating a new legal status quo. There are ways of building generative models that respect intellectual property.”
Rights Tech also has an article about the ins and outs of AI art and copyright, noting “While the litigation against image generators may be new, the debate over whether works produced by AI systems trained on copyrighted works should be considered derivative works under copyright law, and for which a license therefore should have been obtained, is not new.”
There’s also another class action lawsuit where Microsoft, GitHub, and OpenAI are being sued around GitHub Co-pilot being “trained on public repositories of code scraped from the web, many of which are published with licenses that require anyone reusing the code to credit its creators. Copilot has been found to regurgitate long sections of licensed code without providing credit — prompting this lawsuit that accuses the companies of violating copyright law on a massive scale.” The Verge
Some people have emailed me suggesting that these cases will end generative AI before it really gets going, but while these cases are interesting and important, none are intended to shut down generative AI tools.
They are intended to ensure creators and owners of intellectual property are fairly compensated for training models. So regardless of the outcome, AI development will continue apace. The genie is truly out of the bottle.
And it’s not just businesses that want AI development to continue.
I was part of a submission to the UK government on AI and copyright last year along with the Alliance of Independent Authors, and in reading the associated documents, there was a clear attempt to balance reward for creative work with the need to encourage AI innovation, not stifle it.
The US government would presumably be even more keen on ensuring the continuing development of AI. After all, do you really think the US wants another country to be predominant in AI?
Read AI SuperPowers by Kai-Fu Lee if you want a glimpse of the stakes.
Personally, I think the court cases will probably result in some kind of settlement fund for those artists who can prove damages, a change in copyright law to allow for licensing to train models — or it might even be made part of fair use — and there will be an opt-out for anyone who doesn’t want their art/words/music/whatever to be used in training data or used as a prompt.
Fine-tuning models to specific requirements will become much more accessible.
I wrote about this in my book on Artificial Intelligence, Blockchain, and Virtual Worlds, where I proposed a new license to train models with our work in copyright where the original creator gets paid, which would definitely fit with what is being discussed.
For example, a group of action-adventure thriller writers could get together and train a model with our work and then license it to other writers. I’m up for that as I discussed with Charlene Putney on episode 660.
So AI is not going away. It will only become more pervasive in every industry.
Andrew Ng, professor at Stanford, also worked at Google Brain, Baidu AI, and co-founder of Coursera, said that “AI is the new electricity,” in that it will be incorporated into every industry, and the creative sphere is just one tiny part of the transformation.
He also has a great free course on AI for Everyone, which I did a while back, and I recommend if you want to learn more.
In fact, you are also already using AI tools every day — in your car through Google or Apple Maps, if you edit with tools like Grammarly or ProWritingAid, if you publish or shop on Amazon, if you use advertising through Facebook or Amazon, if you use social media on sites like TikTok, YouTube, or Twitter.
Microsoft has also licensed OpenAI’s models for use in their platform and products so these tools are already being incorporated into things you use, and will likely be in MS Word soon, if it’s not already.
You are already AI-assisted. The question is how much do you want to use AI tools in your creative and business practices, as well as in your personal life and work?
So if you want to keep an open mind and try some of the tools, here are my thoughts on the best way to approach them.
Be curious. Be playful. Experiment.
As Oliver and I discuss in the upcoming interview, we are both having a lot of fun with these tools, both image and text tools. They are like a jumpstart for your brain, a crazy co-writer, or co-creator. They spark new ideas and make far more things possible. For the curious creative mind, they are expansive tools.
If you keep a fun, positive, open mindset when you approach AI tools, you may find possibilities you never expected.
As Kevin Kelly says in The Inevitable: Understanding the 12 Technological Forces That Will Shape Our Future.
“This is not a race against the machines. If we race against them, we lose. This is a race with the machines. You’ll be paid in the future based on how well you work with robots … It is inevitable.”
Use AI tools ethically and responsibly
I recommend reading and applying the Alliance of Independent Authors Ethical Guidelines for use of AI.
Some of the key points are:
Use the tools to enhance your creative work, not to infringe on other people’s IP or pass work off as others.
For example, I use Sudowrite (based on GPT3) as a way to help me write better sensory description in my stories, but I do not ask it to “rewrite this text in the style of Stephen King.”
As much as I love King’s books, I’m J.F. Penn and I have my voice and my own stories to tell. I want to use AI tools for my creative vision and enhance my own voice.
I don’t want to create in the style or tone or voice of someone else — and that is true of all artists in any creative niche. Yes, we read other people’s work, but ultimately we want to create our own thing.
If I use any text from Sudowrite or ChatGPT, I edit all generated text to fit my voice and also check it with ProWritingAid’s plagiarism checker, just in case I have mistakenly used someone else’s words. I did this prior to using AI tools anyway, as it’s good practice when you do a lot of research from other books.
In the same way, you can ask AI tools like Midjourney to create images without using the name of a living artist as a prompt. You can get plenty of amazing images without infringing on someone else’s art.
For example, the image on this post is made on Midjourney with the prompt “a robot holding an artists palette, metallic, colorful, detailed.”
Then I checked the final image on Google Image Search to make sure it didn’t resemble someone else’s art.
Be honest and upfront about AI usage
I include a statement of AI usage at the back of my books and I label my AI-narrated audiobooks as well as when I use AI images on my blog posts or in my books.
I have a longer blog post coming in late Feb when I publish my first story with an AI-generated cover, which has also been edited by my cover designer.
How will AI tools impact our business models as authors (and artists)
As ever, it’s about creating original work that you want to create, with your voice, your creative direction, your emotion, and your life experience.
There will always be people who will use the latest tool to cheat and scam and make a quick buck and generally do all the things humans have always done. Plagiarism and piracy are nothing new, so just get on with creating and connecting with your readers and listeners.
Double down on being human, as complex and flawed as you are, and focus on connecting with other humans — however you choose to create — whether that’s with AI tools or without them.
This is one of the reasons why I am so excited about Kickstarter and selling direct on Shopify and other stores. It means we can do amazing quality products for our audience, sell directly and control the platform, as well as connect with the people who buy.
It also means I don’t have to compete on the same platform as all the other creators, as well as the scammers who only target the big platforms.
I can carve out my own little corner of the internet and create and sell happily over here!
(And yes, as this goes out, my Kickstarter for Pilgrimage is running until 5 Feb 2023!)
Even if AI tools can at some point write an amazing novel, or generate a moving song, or create incredible art — it doesn’t matter.
There are already billions of images, songs, and books that anyone can read.
We already have abundance and there will only be more of it to come.
The important thing is creating unique and wonderful stories/art/books/music and connecting with an audience of humans who it resonates with.
It’s about your voice, your angle on the story or topic, your personal experience, your face, your voice.
Your personal brand as a creator and controlling your own intellectual property and your platform is more important than ever in this age of AI.
Circling back to my 2019 article on how AI would disrupt the industry, I also said,
“Humans are innately creative and in this new AI-powered world, we can create even more than we ever dreamed possible.”
I’m more excited than ever about it, and as you’ll hear in the interview, Oliver is too, and we’ll get into that interview in a minute.
If you’d like to learn more, check out Artificial Intelligence, Blockchain and Virtual Worlds, which has chapters on writing in the age of AI and copyright law, written in 2020, but only now really coming true. As ever, I am early …
I also have a course on The AI-Assisted Author which goes into a lot more detail around attitudes to AI, problems, biases, ethical usage, and more.
You can get 50% off the ebook, audiobook, and the course if you buy direct. Use coupon FEB23 at checkout, links in the show notes, valid until March 2023.
There are a ton of newsletters and podcasts out there on these topics, and I wade through a lot of them every week!
I recommend The Algorithmic Bridge by Alberto Romero and Ben’s Bites in particular.
For AI-related podcasts, Hard Fork, Grey Matter with Reid Hoffman, A16Z, and Mindsets and Moonshots.
And check out my backlist podcast episodes and book list at https://www.thecreativepenn.com/future/
Right, let’s get into the interview.
Transcript of Interview with Oliver Altair
Joanna: Oliver Altair is a dark fantasy author, a digital artist, and creator of the Ravensfield Collection. So welcome, Oliver.
Oliver: Hey, Joanna. Thank you so much for having me.
Joanna: Oh, I'm excited to talk to you today. So first up —
Tell us just a bit more about you and how you got into writing and digital art.
Oliver: For me, it all goes way back to when I was a little boy. Creativity was very much encouraged in my household. Both of my parents have brilliant creative minds, my father is an artist. So for me, pretty much I spent all my time writing my stories and illustrating my own stories.
Then when I got my first computer, I got like intrigued about digital art. So I carried that with me until my adult life. And professionally, I became a graphic artist pretty much right after college. So I've been doing that for around 15 years. And when it comes to publishing, I wasn't aware of the self-publishing world until 2016/2017, so I was a bit late to the game. 2017 is when I published my first dark fantasy novel.
Joanna: Okay, well, no one is late. So you certainly weren't late then. And anyone starting now, you are not late. I mean, I feel like I'm a bit with Amazon on this, every day is day one. There's always something new. And we'll be talking about new stuff today.
Let's just tell people where you are in the world in case they're wondering what your accent is.
It's only very slight, but I think people will be interested.
Oliver: Yes. So I'm originally from Spain. I was born in the southwest of Spain, but then I moved to the United States in my early 20s, so I was there for pretty much until I was in my mid to late 30s. So it's been a while, but now I live between the US and Europe with my husband. Right now, today. I'm actually in the Lisbon area in Portugal, which is quite lovely.
Joanna: Oh, lovely place.
Tell us more about the Ravensfield Collection. What is this? And why did you create it?
Oliver: Yes, so I called the Ravensfield Collection my own museum of the weird. So it's an imaginary museum that I created and it has pieces of artwork that go from oil paintings to sculptures to just like objects of art, pretty much, whatever you could find in an eclectic museum. Think for example, the Metropolitan Museum in New York City.
Then each little object has a short story, a bit of text related to the object. And it's usually something in the realm of like weird fiction, which is what I write about. And some of the stories are actually interrelated, so they give you information about that world, about the family that has curated the collection, etc.
The one thing that excites me the most is like all of the art pieces I made using an artificial intelligence image generator, which I believe adds a little bit of like weirdness and mysticism, if you will, to the whole project.
About the “why” of I decided to create this, I think there was both a creative reason and also a more practical reason.
Creatively, I've always loved anthology projects, things like, you know, The Martian Chronicles, or Chuck Palahniuk's Haunted, or of course, The Twilight Zone. So I decided that this was my chance to create my own.
Then at the same time, when I started my publishing journey, I was actually pretty happy with the response of the readers that I could find so far. But I think I felt a bit short when it came to widening my audience a little. And probably some of your listeners will share this feeling with me.
If you write something that is a bit hard to position on the bookshelf, it gets really tricky to find the correct readers.
So I thought this was a great chance to create a project that would interest the minds that I'm trying to reach, people that are just interested in things that have are a bit out there, a bit weird, a bit in the realms of strange fiction and horror and all that kind of stuff.
Joanna: And actually, this is why I'm talking to you. I mean, I've had a lot of pitches around, “let's talk about AI art.” And then I went on your Ravensfield Collection, and I was like, ‘oh, we have a similar mind in that way.' In that, you know, I've got Guillermo del Toro's Cabinet of Curiosities, the book about his house, which is just full of the weirdest stuff. And I have so many very expensive hardback books that are full of weird pictures from weird museums.
When I saw yours, I was like, ‘ah, this is a great idea.' And it totally gets into the things that we maybe can't find pictures of either. This is what's so awesome about AI art. And I might have to do something similar for my book Desecration, where just there are no pictures of some of the things I'm writing about. So I love that you're doing this from a creative point of view.
You mentioned the readers' response, so what has been the readers' response?
Have people been like, “oh, I want to know the story behind that object”? Has it actually led to book sales? Can you tell?
Oliver: So when it comes to book sales, right now, I'm in the middle of a huge project because I decided that the story that I wrote in my first few books, I felt the scope was too small. So I decided to actually re-edit both volumes, which is also one of the reasons I decided to put some effort into Ravensfield because that's a project that is going to take me a long time, and I wanted to still be active and be sharing content with readers out there.
The response so far has been pretty fantastic. Like people are just very intrigued and also I like surprising people that fall into the Ravensfield Collection because I think the first reaction is a bit confused. Like they don't know what the thing is about, is this a real place.
Then they start investigating and digging, and they get more and more into it. And then they message me, and they say, I cannot wait for the next object to drop. Or they actually want to meet the people in the family, or they request that I do images that are actually the portraits of these characters that I'm talking about.
So it's opening a lot of really fun interactions with people. And also, it opens the conversations in regards to AI art because that's also something that most people didn't expect.
Because to be fair, most people don't know that AI art even exists.
So they question like, where did I find those objects, if it's something that was like laying around my house and I decided to write a story about it. And I just have so much fun because like being in that kind of in-between ties so beautifully with the things that I write about.
Joanna: And I'm so glad you use the word “fun” there because we're going to get into some serious issues and problems that people have.
But when you're creating with these tools, and when I'm creating these tools, I'm having fun.
Just before this, I was in Midjourney, I was making some more character pictures, I was making some background pictures, and I love what you're doing there in terms of like, it's fun, and then you're using it to world build.
I feel like maybe as professionals, we get to the point of thinking everything we do has to generate income, like that particular thing has to be revenue generating. This almost goes back to what is creatively fun and just part of just the joy, I guess, of creation, which almost takes us back to the beginning of why we did this in the first place.
Oliver: Oh, absolutely. I couldn't agree more. And also, I probably am not alone in what happened during COVID, like a lot of us felt very creatively drained.
And for me, when I discovered this whole new world of AI art, it was like some spark was reignited, in a sense.
Joanna: Yeah, I think I was the same. COVID just was not good. It was not good in many ways. But creatively, it was just dead. So I love this. Let's get into some more specifics.
Which AI image tools are you using? What other tools are you using as an AI-augmented creative?
Oliver: Yes, so I use Midjourney. It's the one that you just mentioned. There are many apps and software when it comes to image generation out there, and I feel like every time I close my eyes there is new one. And they get more and more specific, which is pretty fantastic.
Right now, I decided to master just the one and then probably I'll move on because I'm the kind of creative that likes to have like a hefty toolbox, if possible. This is the one that has given me the best results so far for the goals that I have in mind.
Also, like you talked with Derek Murphy about this, so I will tell your listeners to go back and listen to that one because that was a great conversation. And he goes a bit more into depth about the nuances between the different tools.
Joanna: So well, let's talk about how you use Midjourney then. I'll tell you how I use it because I am not graphically trained, I'm not a visual artist. I do think they have some kind of visual taste.
But on Midjourney, like earlier, I did a “portrait photo of a female combat photographer against the backdrop of burning buildings.” So my prompt has a character type and a background, and I'm using version four, so I use the — v 4. And what I get out of that is pretty blooming good, to be honest. But you're an artist, you're a graphic artist —
What are your prompts like when you're prompting these engines?
Oliver: So my prompting varies depending on what I'm trying to achieve. What I do a lot is like — so I would say I like that you brought the word “taste”. There is definitely a taste level that you need to have to get your results to where you want them to be.
And also, for me, it's not only taste, it's also a tool that requires a lot of curiosity, and it requires you to be very observant. You need to have a very wide repertoire of references and words that you can use to talk to the machine, to the algorithm. And machines are very precise.
So in a sense, it's also like talking a new language from scratch because you cannot talk to Midjourney as we are talking, you and I, because it wouldn't understand you.
And for me, you need to find a balance between the result that you have in your head, and then you need to also leave the algorithm a bit of freedom because, usually, the machine is going to surprise you. And in my case, it always surprises you in the right ways, and it opens different pathways and different possibilities that you didn't think about.
In my case, what I love to do is mixing things that in our human logical minds maybe wouldn't mix. So for example, I tried to mix a fashion designer with a building.
And then the computer does it so beautifully, yet so seamlessly that the results are always so so bizarre. And then I take that into pretty much everything. And then depending on like the mood, then I add a lot of extra information when it comes to lighting, when it comes to color palettes, and when it comes to different mediums or textures, things like that.
So the good thing about image generators is the foundation is really accessible. So anyone can go into the tool and start experimenting. And the results, as you said, from the get go is pretty impressive. And then you can iterate and grow from there. But pretty much you have results that are shareable from the very first hours that you're using Midjourney.
Joanna: You've raised such great points there. And again, I just want to emphasize to people, when you're prompting, you mentioned a repertoire of references, so fashion designers and buildings. So you're not saying ‘house' and ‘fashion designer,' you're saying specific names around fashion designers and buildings.
And then you said lighting, color palette, textures. And the difference in a prompt, like if people haven't seen the difference in prompt — so my husband who's not at all visually artistic, the other day he did a prompt, and it was like ‘cat samurai.' And, I mean, it's a cat samurai, and it has a nice cartoon of a cat.
Oliver: Yeah, that's pretty cute.
Joanna: Yeah, it is cute, right? It's cute. And the picture that comes up in everyone's head right now is probably enough, but that would never be enough for you. Like you would never put cat samurai, right? You would come up with all kinds of other things.
I actually screenshot an artist doing a prompt earlier, and his prompt was like 500 words. You know, so I mean that's the difference, isn't it?
When people say, “oh, there's no room for real artists anymore” — but prompting and using the tools is a real art in itself.
Oliver: It is. I mean, tool is the keyword for me.
Midjourney, it's a tool, like a pencil is a tool. It's not a living being that is trying to take over the world or trying to obliterate artists from the face of the earth.
It is a tool that might interest some artists, it might not interest some others. And then it has a learning curve. And then it has a lot of experimenting that you can do.
The thing is, like as any other thing, it requires some extra thought if you want to find your own voice. It's pretty much very close to writing, right?
Like when you start writing, you're a bit shy, and you try to stick to like a lot of conventions, and then you start breaking free, and then you use things that are a bit more daring. To me, it's pretty much the same mental process.
First, you are a bit shy and you do prompts that are a bit less complex. And then when you start finding your footing, you start getting a bit more brave and then a bit more complex.
And then your prompts are larger, which by the way, not always means better. Some people confuse and they think that if you write pretty much the Bible as a prompt, you're going to get a better result, which is not always the case. It's like anything else, it depends on what you're trying to achieve.
For me, it's a very, very beautiful creative process because it requires a lot of mental clarity.
You need to have a very clear picture in your mind of what you want to achieve. And then you need to translate that into a prompt that is useful to the machine.
So just that is going to challenge your comfort zone creatively time and again, and I think it's just a very, very good way to keep your mind sharp.
Joanna: And so you mentioned ‘voice' there, you mentioned a clear picture of what you want to achieve. We've mentioned taste.
I always talk about creative direction, you know, and I talk about having maybe a supercar sitting there outside your house, and the supercar is not just going to do stuff on its own. You have to get in and drive it and take it to where you want it to go, and you're the one in charge.
So I think that this is so important, these tools don't have any agency.
I mean, as we're recording this certainly, in January 2023, there is no sentient AI artists that's like, “oh, this is my AI voice. This is what I'm going to create.”
And so everything that people are creating is all through the lens of what they want to do. And it's a bit like you can give the same writer a writing prompt, and they will create completely different stories.
You and I could have the same idea of something we want to create, but our outputs will be completely different using the same tool.
It is a tool. And obviously, we agree on this, but people listening, I think are struggling with this idea.
Oliver: Yes, absolutely. And I think there is a whole misconception about what it is and what it does.
And I think some people think that when you prompt, pretty much what you're doing is like a Google image search or you're looking for an image on a stock service. Like they think that what the service does is finds an image that kind of matches what you're saying and either gives you that same image or a carbon copy. And that's not even close to being remotely true.
You need to think that these tools are trained in billions of images around the world. And what they do is they understand what you're trying to achieve. And they take bits and pieces of all of those images of this information in pixels, and then they give you a series of results that you can each rate.
You can then create a whole new image that is going to be completely your own, and it's going to show your personality.
Especially, I think it happens to all creators, at some point, things that look too formulaic are not going to be good enough for you because we are very anal when it comes to finding our voices.
So that's when I think just by like the nature of being creative, you'll try to push new boundaries.
Joanna: Hmm. Yeah, absolutely. I've got some sort of key points of what people are objecting to.
First of all, “is AI art really art?”
And I mean, my opinion, first up, is there is no true AI art because humans are driving it all.
Also, I think there's a very clear difference between like — and I see it on the Midjourney open stream all the time, people upload a selfie, and then they say, ‘make me into a Pixar character.' This is a common thing.
Or make me into a Disney character or something like that. And there is a very, like Pixar, specifically, a very key visual cue that is a Pixar image and people's selfies as Pixar characters. That is not art. I mean, that's just kind of fun. It's what people want to do.
But there's a big difference between someone like yourself or people who are actually creating art with these tools, or people who are just using it for fun. So what are your thoughts on this? Is AI art really art?
Oliver: Yeah, I mean, I find it really funny because, is X, Y and Z art or not art? It's pretty much a debate that has been happening since art exists. So it's not new.
And also I think it's pretty pointless. I mean, is it? Isn't it? I mean, who knows, depends on who you ask. The things that are getting confused here is the novelty of new “toy”, and people that see potential creatively or professionally. Being so new, and I find it really fascinating, some people online are already treating AI art as if it were a mature industry. It's not, by any means. It's still in very, very early stages.
So it's, of course, like very common that people see a new and flashy thing, and they want to do silly experiments. And that's totally fine, and they can have their fun.
But I think all of those people eventually are going to find their interest somewhere else, and they're going to move on. And then the people that are going to remain are people that see a potential beyond the novelty. You know what I mean?
Joanna: Yeah, and I mean, photography is the classic example that people are using, which is when photography came along, early photography, all the painters went, “What are you doing? That's not real art.”
And then people started taking photos. And then it went from the old film to digital. And then there's all the tools like Photoshop you can use to manipulate it. And so on every scale — I mean, music, you don't need to play a violin now to make violin music, you know, all of these things have tools.
Yes, and it's funny you say that about people getting bored. Like, again, like my husband, for example, he does like occasionally doing like the cat samurai stuff, but he has no reason to use AI visual art in his daily life. And so he will rapidly get bored with it. And you will carry on, I will carry on in a lesser way than you will, but we all try these things and then find a place for them. And if there is no place for them, that's also fine.
Oliver: Absolutely. And I think finding the place, that's what people need to think about.
If it's a tool that has potential — which I believe AI does for several reasons — it will find its rightful place and then people will use it accordingly.
Not only that, like AI art existing probably is also going to open new synergies between AI and other artworks that already exist, in the movie world or even like a traditional name in writing.
When photography came to be, or film, also the other arts caught up and created new things that wouldn't exist otherwise.
And I think this is going to be pretty much the same thing. I mean, sometimes I wonder if the difference is the speed, right? This transition is happening so fast that sometimes I think the fear is related to this kind of like vertigo effect because things are just changing so rapidly that you cannot even grasp what they mean, or what the change might be, or if it's good or bad, and then people get just like absolutely stunned.
Joanna: Yes, and that is why I like to keep coming back to this because a lot of people actually have emailed me about this show because I've been talking about AI since 2016 on the show. And people are like, oh, I feel a bit more comfortable because I've listened to you talk about it over the years. So hopefully, this conversation will help more people who need to adjust more.
Again, you mentioned film there. I mean, the gaming industry has been using AI generation for a long time. And we're only going to see more of it.
Let's just talk about a couple of other issues that people bring up as regards to AI art. So first of all, you did mention that these models are trained with images. And this is one of the sort of things people say, AI art is stealing from creators, they're trained with images that were copyright protected, this should not have happened.
What are your thoughts on the AI tools “stealing from creators” by training the models?
Oliver: Yes, like, listen, there are a lot of artists that are turning this into such a huge controversy and pretty much into a war of us versus them, and I'm not going to be that person.
I think it's natural that creators have concerns about their property. And at the end of the day, it's up to each creator to decide how they want to use their artwork and put it out in the world and license it or not license it. So I understand it's an issue that requires some thought.
The problem is like the industry being so quick and so new, there are a lot of question marks. So my problem with this kind of mentality is I think there is a very thin line between being watchful about your property and being egotistical and petty. I think it's important, as a creator, to try to put generosity out there.
Some people are very close to the idea of someone mimicking their style. To me, personally, that's a compliment. If someone wants to mimic your style, I would guess to learn, or because they think you're so wonderful that they want to be you, take the compliment. That said, there are always going to be bad actors out there.
There is a difference between a person using you as a reference to grow and learn and someone that is trying to pass their work as their own and profiting from it.
Joanna: Yeah, I mean, that passing off, it's very difficult because these tools don't technically plagiarize, as in the image is not exactly the same image, it will have differences, whatever it is.
But like you said, some of the tools are now allowing artists to remove their names from prompts. [Ars Technica on Stability AI]
It's funny, I mean, again, I agree with you totally, I think people have a choice to use it or not, but personally, I think if you get your name removed from what may be whatever the next form of some kind of visual internet, then isn't that going to hurt you in the long run. Do you then become invisible?
So I've been thinking about this a long time, and I also feel like I have utilitarianism, where the greater benefit to the whole is more important. Obviously, I have a lot of intellectual property in words, and I know it's all in the models. Of course, it's all in the models. All of our work is available somehow online. And it's like, well, again, if people want to train their word models with J.F. Penn or Joanna Penn, then isn't that a good thing?
This is where I think personal brand and connection with readers and viewers is so important.
I control my relationship with my audience.
And so that, I think will become even more important as we go on. It's like, how can you prove to your audience that you're human?
Even though you use all these AI tools, they're connecting with you, the human behind that art.
So I'm not particularly worried about that myself, either.
Oliver: Yeah, also, I think this is an issue that as soon as the dust settles a little, it's going to disappear. Things are very turmoil-y, because we are in a moment that we are not there yet. But as soon as there is a better understanding of these tools, I think these fears are going to just naturally disappear.
So for me, it's just important to remember, as you were saying, as a creator, like an AI tool, it's not a clone of your head. So you can mimic a person all you want, but at the end of the day, the creator is the creator.
As a creator, you need to be confident that the content that you're putting out there is coming from your mind, and that's what is creating the connection with your audience.
You need to remember, if you're doing content that is connecting, and it's causing a reaction, and your audience is reacting favorably to it, they don't care if you use AI to make it, or if it's an old painting, or if you're writing with a goose feather or with a typewriter.
What they like is that connection that you're putting effort into nurturing. And that's something that AI cannot copy.
Joanna: Yeah, I do think this is more and more important over time. So we've talked there about the inputs into the tool. So the training data that's gone into the tools, clearly, some of that is copyrighted.
But the output is another interesting thing, and there is no legal stand on this right now. Although again, as you say, I think this year, 2023, things will start shaking out in terms of the legal side.
Midjourney, and I went and checked again, I've read their terms of service, and they do basically say that what comes out of Midjourney is yours, the creators, but that people can remix it, they can do whatever. If you're creating in public, which I do, people can just do that.
You've put your images under Creative Commons Attribution-NonCommercial-ShareAlike.
So can you talk about that? What is that? And why did you decide to go that way?
Oliver: Yes, so Creative Commons is a nonprofit that helps create or make these licenses that pretty much say, it's a way that you're telling the whole world that you welcome people to interact with your content, but then there is some conditions that need to be met.
This is especially useful if you're a creator that puts a lot of content online. So in my case, because also I want to monetize my content in the future, I made this license that it's a bit more restrictive.
I'll go one by one, so people can understand. When you say attribution, that means that everybody is welcome to use my art, to share it or to do derivative work, but they have to let the world know that it comes from me. So they have to mention your name.
When you say that something is noncommercial, that just means that they cannot profit from that piece of work. And when you say it is sharealike, it means that they need to use that same license, use those same conditions, with their new art piece.
So this would be pretty much as if you're granting the world a fan art service, in a way. You're just encouraging people to share your work, to have fun with it, but you don't want them to take your money or your credit from you.
When it comes to Creative Commons, what I like the most is it's a way to encourage both that generosity that I was talking about, and also network effect. You're telling creators and people around the world, please play with my material, but also you're giving yourself a little bit of protection because if people are completely disrespecting your licenses, you can talk to them and say, “Listen, like, I'm very glad that you're playing with my artwork, but you need to follow these steps.”
Those licenses are also very interesting if you're trying to build something like a multiverse these days because you can remove the noncommercial and then what you're telling people is, this is my world, and then you're free to play with it, and you can profit, but also you need to share it alike, meaning like there are certain conditions that need to be met.
It's just another way that creators use to pretty much like state how they want people to react to the world, to interact with your material, and grant those extra licenses and be a bit more clear about what your intention is online.
Joanna: Yeah, and I certainly think copyright law in general is going to be much discussed—
Oliver: Oh my goodness.
Joanna: I know, it's very, very difficult to really know what's going on. And I have the same kind of attitude, which is I'm treating the pictures I create on Midjourney as essentially Creative Commons. And I'm still doing it in public on the Midjourney newbie stream, and then other people can just take it and do whatever they want with it. And that doesn't bother me, but art is not my business.
So this is interesting because you are a graphic artist in that you also have a job this way. So if people listening, if there are people who are artists who want to make money and use AI in their arts — and I mean we've got things like Adobe and other commonly used tools are all incorporating AI now into the design tools.
So is it just a line between, say, if you generate an image with Midjourney, but then you use Photoshop to edit that and change things, then that's copyright and then you can sell it?
What do you think about where the line is around getting paid for AI images?
Oliver: So that's the thing, you just raise a very good point. What happens when you put an image in Photoshop? Can you copyright it right now, and then monetize it? I wish things were a bit less blurry in regards to that right now.
The only thing I would say is monetization, when it comes to artwork, is a decision that creators need to do before they put content online. Because if not, it's going to be very messy if you have to make that decision afterwards.
So you need to be very clear about what your goals are, and what part of your artwork is going to be monetizable, and what you're willing to share for free, and then make the decisions accordingly. Because then that's going to help you decide the service that you want to use, and then what licenses that you're going to look for, and then you can build from there.
I wouldn't recommend just starting putting tons and tons and tons of content online and then getting surprises about someone using it and profiting because you weren't clear enough.
When it comes to copyright, I think from my understanding what is happening today, and again, like as you said, it is such a mess, because I mean, copyright is messy no matter what, and AI is not making it any easier.
I think the difference lies in like when you can demonstrate that you are using this as a tool and there is a lot of human work behind it. So it's also very tricky to prove that, but some creators are actually going through that route and talking to the registry office and saying, “Listen, like this is my project. This is how I used an AI tool. And this is the steps that I took, etc, etc.” And sometimes they get the registration and nothing happens.
So, I mean, I don't know. I mean, I tell people just be careful and be watchful out there. Make sure that you're making the right decision so that you're protecting your company and your property enough, but don't be losing sleep over it every night.
So some point there is like nothing that you can do. Just be flexible and be willing to do as much as you can, and pretty much that's it. Also, seriously, I don't think there is anything else. And maybe I'm missing something, but I don't think there is anything else you can do, at least at this stage.
Joanna: I mean, I feel like this is nothing new either. I've been putting stuff online now since 2008.
Every single blog post, podcast, video, book, everything I have ever made has been plagiarized, put up on another site or pirated (and by humans, not AI.)
And sometimes I get notified straightaway. I'll get pingbacks, on like this interview, I will get pingbacks from websites that are just scraping and posting. Sometimes I will go about trying to take them down, and a lot of the times I'm just like, oh, my goodness, I just can't keep up with it all.
That's why I emphasize the relationship with my audience because the people listening to this or reading the transcript here with us, they're our audience.
And I'm just not going to worry about the rest of it because scammers and spammers, and whatever, will never stop.
AI will supercharge it for sure, but if there are billions — well, there already are millions and millions of books and millions of millions of art out in the world, so it still comes down to the connection between the creator and the consumer, the reader, the viewer. And that's something that we have to find as individual creators, I guess.
And the AI is not going to do that. And the people who just scam other people's work or plagiarize other people's work, steal other people's work, they are not going about it in the same way. They’re not building that relationship.
Oliver: Oh, of course not. Of course not. And at the end of the day, it's great that you're pointing this, Joanna, because like I think there is something that also creators tend to misunderstand.
You cannot think about monetizing when you have nothing to monetize, which means that you need an audience.
You cannot like start selling products left and right if you have no one to sell those products to.
I think the mistake that a lot of people get into, and you mentioned this before, and it's because like it's really easy for us, as human beings, to get very caught into that making a profit mentality, and then we forget that that depends, as creatives, completely on having a healthy audience and interest with you. If you don't have it, you can forget about it. And that's the way it is.
Scammers don't care because they're going to use whatever is available, and then they're going to move on.
One thing that I keep telling people that also ties to what you just said, is like if you're lucky, you might get a ping and you might find out. Most likely, you will never find out. So you might as well just go and enjoy your day.
Joanna: Absolutely. Now, you mentioned earlier about this sort of war going on in the community. And I mean, this is starting to happen in the writing community too, as people find out about AI writing tools, which we might come back to in a minute.
But in terms of what people think, you mentioned earlier right at the beginning that your father is an artist. What does your father think about this?
How do you talk to people that you want to convince that what you're doing is real?
Oliver: Oh, not at all. My father loved it. I mean, he's not interested himself. He's also in his 70s, and he mostly does oil painting and these very intricate ink drawings.
And he's not interested in using it himself, but when I explained what it does, and I showed him a lot of examples, like he was just intrigued and excited and he saw a lot of potential. He encouraged me to just keep exploring and be better at it. Like for him, like he didn't see it as like a threat or anything else. Quite the opposite.
So yeah, the conversation, for me, Joanna, went very well, so I would encourage you to maybe like open it and see what happens. I think, I mean, from my perspective, like this conversation goes wrong when you're talking with people that lack a bit of confidence as creators. And that can go both ways, it can be like both the AI artist and the traditional artists. So it becomes a battle of like, who's better? And that's a mess because that's not what the discussion is. Like, you don't need to be convincing me that you're better than me, that's absolutely pointless.
So I think it's just like, I don't know, for me this is like everything else in the online world. Just be gracious, be understanding and listen to other people. And instead of shrieking and insulting, because that's already happening. So that happened really fast in AI art, there's already like this battle of insults, which, by the way, are very creative. So I would encourage those people to put that effort into doing images.
For me, it's just about confidence.
If you're a creator, you need to be 100% confident that what you're putting out there is good and is resonating with people.
And if it's not, then it's up to you to be flexible enough to change course and try something new.
You cannot blame a new tool or technology or other person that is doing things better than you for your lack of success. Because the only thing that you're doing is driving yourself completely insane. You're blocking yourself creatively, you're not evolving, and you're just being bitter. And what's the point?
Joanna: That kind of circles back to what we said at the beginning, around the fun side, the creative side. People listening, if you haven't tried any of these tools, please just try them.
I feel like a lot of people criticize before they have tried it. And once you try something, I think the fear goes out of it.
Given that most of our audience are writers more than visual artists, let's just mention a writing tool that they might have tried it. And we were talking about this before the recording, which is ChatGPT or some of the other writing tools. So you've been using this as well —
Tell us what you're using generative writing tools for.
Oliver: So I think ChatGPT, specifically, and probably the other ones out there, are great brainstorming companions. They're not going to write the book for you because I don't think that's a thing, to be honest. I think books have way too many moving pieces for a machine to do it to the level that you're going to ask from it. It's great if you want to bounce ideas, and very quickly.
Also again, because it's a machine it always surprises you. So I've been using it when I feel a bit blocked or like the direction that I am going is not giving me the results that I want.
Both for fiction and for generating art, I just go to ChatGPT and start asking questions about like, can you give me a synopsis for this? And then I'm like, maybe make it more Gothic, maybe make it more cyberpunk, maybe add this to the mix.
And then those results helped me tremendously to unblock because it's pretty much like you're bouncing ideas back and forth, and back and forth, and back and forth.
Again, really fun, so I highly recommend it. It also pushes your imagination a lot, and also it helps you relax. And it will definitely show you ideas that didn't cross your mind before.
Joanna: Absolutely. And this is why I talk about being AI-augmented, or AI-assisted creatives.
Because again, like you said, it's a tool that helps us think in different directions. And it's like having just a ton of different brains that you can access that spark ideas for your own creative journey.
I love that you mentioned creative confidence, because again, even if you can just click a button and output a perfect image or click a button and output a perfect book first go, that doesn't matter.
Because it comes back to what we've talked about, it's what you want to create, and then it's about your connection with other humans.
Like we said, there are already too many books in the world, there are too many images in the world for any person to ever read or look at or listen to music. So there will just be more and more and more content. But that's not the thing, right?
We're going to keep creating because we are creative people, and this is just going to expand the possibilities.
It just comes down to what do you want to create and who do you want to connect with.
Oliver: Yes, absolutely. And for me, it's just a beautiful exercise of the imagination. I think if you are an author, we also have an advantage that is that we live in the world of ideas pretty much 24/7. So now we have a tool that makes those ideas manifest into the world really quickly. So for me, that's just a huge advantage.
You need to think that like, specifically with images, it's limitless. So you can go as small as designing a pebble on a beach, and then grow all the way to designing the map of your whole universe, and everything in between. To me, that just fills me with so much excitement because I see so much potential.
Also, it opens the realms of possibilities for people that maybe are afraid that their ideas are too out there. You know, we all share that fear. I think it happens to all of us.
Sometimes you write something and you think, oh, this is just too strange, this is not going anywhere. And then, well, maybe it will. Maybe you just need to explore it a bit further.
And I think those tools just push you to take that journey to do more exploring, to find new ways to express yourself. Just for the mere fact that that's a thing, I would just recommend people go ahead and try them. And then if they're not interested, that's absolutely fine, but I think they're just going to see a lot of potential just like we did.
Joanna: I'm so glad you're excited. I am too.
So where can people find you and everything you do online?
Oliver: So the best way to connect with me is through my website. That's OliverAltair.com. There you can find a contact form, so please, anyone that has a question about AI or fiction or just wants to say hello, feel free to shoot me a message. Then from there, you can also find the links to my social media and to the Ravensfield Collection.
Joanna: Brilliant. Well, thanks so much for your time, Oliver. That was great.
Oliver: Thank you so much, Joanna.
Wes McBride says
There are some unconscious prejudices we should be wary of when talking about AI. First is the idea that AI developers don’t care about consequences (break things, go fast). Second is the idea that artists are luddites (except for the few enlightened ones using AI). I’m sure there are other prejudices, but I’d like to address the second.
“Yeah, and I mean, photography is the classic example that people are using, which is when photography came along, early photography, all the painters went, ‘What are you doing? That’s not real art.'”
This classic example is not based on actual history or fact. It is a popular straw-man used in discussions about AI Art. A false idea set up to make people who are concerned about AI look unreasonable or fearful.
This example is so common online that many people take it as fact without question. It sounds like something that might have happened. Except that it didn’t. Not in the way many assume. And why would people know the facts about this? Few people study Art History.
There isn’t a lot of evidence that painters resisted early photography. Or that artists have ever been naturally inclined to resist technology. Actual evidence points in the other direction. Artists have generally been of the forefront of adopting new tools and materials.
Daguerre himself was a painter. As was Samuel Morse, the man who helped popularize photography in the United States. As was Sir Charles Eastlake, president of the Royal Academy during the early years of photography. He co-authored an influential paper suggesting that photography could easily be used as an artistic tool. Many of the impressionist painters were avid admirers of photography. Many nineteenth century sculptors and painters used photography as a tool to understand animal locomotion. And so on…
The idea that artists opposed photography is a myth.
Some art critics (mostly not visual artists themselves) did not like photography. Baudelaire comes to mind. Also Ruskin.
Perhaps the false idea that artists are opposed to technology comes from Ruskin? He is one of the few art theorists non-artists read in school. Ruskin’s ideas were largely in reaction to mass production and automation of craft industries, not photography. Ruskin was so extreme in his views that he envisioned his much younger wife as an untouchable object d’art–and treated her abysmally when she didn’t live up to his inhuman ideals. Many painters publicly distanced themselves from Ruskin, including his own protege, when this abuse became public.
Of course, there were shows in the nineteenth century that excluded photography. There were also shows that excluded painting, sculpture, and particular artistic styles. The Academie des Beaux-Arts was behind the times on photography, but the academy was also behind the times on almost every artistic style or movement after Romanticism. Not to mention the academy was set up expressly as a school for painting and sculpture (not as a general arts school). Bouguereau, the poster-boy of the academy if ever there was one, was an avid collector and an admirer of photography.
Twentieth century golden age Illustrators did grumble about photography during the sixty-year period over which photographs gradually replaced illustrations in newspapers and magazines. But that was a complaint against job loss, not photography per se. Most illustrators of this time used photography to help create their illustrations.
By that time, photography was nearing 100 years old and entering a golden age of its own with photographers like Strand and Stieglitz taking inspiration from paintings and appearing in shows alongside Cezanne and Picasso.
Artists generally aren’t the ones resisting change.
Very interesting article.
To compare Ai with the invention of the camera is not a fair comparison.
Ai is game changer like no other and we’re only in the foetal stage.
This Ai is going to change every aspect of our lives and work.
It’s not some warm and fuzzy little app.
I’m aware of the submissions to the U.K. Parliament. It would be interesting to know how many of them read the full submissions or even the top briefing sheet.
This will destroy the lives of writers, artists, actors but not just creatives – it will make millions of people (across all areas of industry ) unemployed.
If companies like Getty are worried about Ai – then all creatives should be worried about this.
This Ai thing is much bigger than a new little tool that can pretty up our art work or help us write a snappy line.
Personally I think it’s game over.
This Ai thing is going to eat us alive and many people seem to be happy to feed themselves to the machine.
Joanna Penn says
Like all technological shifts, there are positives and negatives. But it is already transforming my creative business in a positive direction, so it all depends on how much you embrace the tools.
Getty are not fighting the technology, they specifically say that in the case, which I outlined in my introduction. None of the law suits are trying to stop the development of AI, they are objecting to the training data and methods.
Getty are trying to put the brake on Ai in a nuanced way.
Getty are fighting the fights that they feel that can win.
I still stand by my assertion that it’s game over.
I’m not being defeatist.
The tech industry knows what’s coming doing the tube and hence the massive lay offs.
I would love to be wrong on this point.
Steaphan Kay says
Here is a mind twister: I started using ChatGPT to create Midjourney prompts of image concepts I have. It works very well and writes much better prompts than I do.