Depending on who you ask, AI is either the greatest advancement in history since the invention of the wheel or a malevolent spectre destined to drag humankind into a techno-authoritarian dystopia. Who you choose to believe probably says more about you than the technology itself. Many in the art world are painfully ignorant about the everchanging developments in the field of artificial intelligence. Some are even proudly ignorant, rejecting it as an affront to the very notion of art itself. However, there exist a storied few who have harnessed the powers of this futuristic apparatus and used it to visualise entire new worlds. One of these is none other than the enigmatic Joann.
Interview taken from METAL Magazine issue 50. Adapted for the online version. Order your copy here.
Joann has built up a wealth of images on Instagram depicting hyperrealist iterations of concepts that once would have been considered almost undepictable. From Parisian maisons entirely of hair to ocean dwelling creatures knitted in technicolour, there are no limits to what the account might post next. Each collection is accompanied by concise yet astute captions, giving us a glimpse into the prompt that may have sparked each of these imagined worlds. Although these cyberspatial absurdities may seem to exist many dimensions away, many link back to brands and establishments from the real world. As well as collaborating with a plethora of haute couture fashion houses – Gucci, Versace, Valentino and so on, Joann has also produced work for a diverse range of companies ranging from Oatly to adidas.
Beyond the artworks themselves, Joann’s digital footprint is relatively modest. The artist, who lives and works in Armenia, has committed the thriving Instagram account almost exclusively to the creative, with the personal remaining a relative mystery. We caught up with Joann herself to delve into the person and the processes behind the psychedelic creations of the Internet’s most in vogue AI artist.
Hello Joann! Your mononymic moniker feels like a good place to start. How did you come to be known as Joann?
Well, the name Joann (pronounced Jo-Ann) actually comes from my adolescence. When I was a teenager, I really loved Joan Jett and also Joey Jordison, the drummer from Slipknot. And my real name was Anne so I decided to rename myself. I just added Joe to it on Facebook and Instagram at that time. And it went like that, and everybody just started to recognise me as Joann and now even new people know me as Joann and they don’t actually know my real name.
Yeah. Well, that’s the thing, your online presence is very – well, there’s not much there, you know. It’s quite hard to find anything about the person behind the art.
I’m just not that social anymore really. About ten years ago maybe, I was very social on all of the different medias. I loved posting photos and everything from my life. But over the years as I grew up, I don’t know, I kind of became antisocial, so I don’t like to share anything from my personal life. To me, that’s just it. Like, besides the arts.
So is your Instagram profile intentionally anonymous, like Banksy, or?
No, no, it’s not intentional. It’s just my personality really. I’m kind of introverted and also, I’m just very lazy when it comes to sharing anything from my life. I just don’t really see the point.
Ok, so in terms of your Instagram, could you just tell me about the creative process for how a work comes into being? How does it all start?
Well, actually I just go to Midjourney. I just type text. People like to say this about AI art. You just type the text and that’s it. It lasts two minutes. But obviously in reality there’s much more to it than that. I have lots of ideas in my mind that I really want to see visually that don’t exist in the real world and Midjourney helps me with this a lot. It helps me to visualise this. A lot of the ideas I play with come from my childhood and all of the years of my life where we didn’t have this possibility to just instantly visualise anything imaginable. I love just playing with it. The skill comes from the experimentation. You have to know the structure of languages. I have a diploma in journalism, and I am actually a linguist. Although my background is in linguistics and journalism, in real life I actually worked as a creative director and a designer and an illustrator for ten years. It’s really mixed up. But then when AI came into our lives, all of my professions really mixed with each other. Crafting the text prompts themselves is a huge part of the process. This is why it’s easier for me to explain to the AI what I want to see rather than other people who don’t have this mixed background.
And do you often find that text prompts you have fed in don’t come out all that well?
Yeah, actually at the start it’s a whole other thing. Whenever I start texting, for example, I have an idea that I want to see. I don’t know, let’s say a tree made of ice cream, for example. I just type in the short text for a tree made of ice cream. And then I see the result. I don’t like the result. The first results are generally bad because at the start the AI doesn’t understand you very well. So then you need to explain with a different bunch of words and different variations of sentences and there is this back and forth process of changing the words all the time. Ok, so if it doesn’t understand this word or I don’t like the result then I’ll delete this word and replace it with a different one and see how it’s interpreted. It really can take a day or even days, hours of altering prompts to get the appropriate result.
And then when you’ve finally got a generated image that you’re pleased with, is there anything that you do to it afterwards?
Of course, I always do the postproduction in Photoshop and often also use other different AI tools for various needs. For example, there are different kinds of upscalers that you can apply to the images with prompts. It’s basically the same image, you just upscale it and make it more beautiful and with a better resolution. Again, this fine tuning can work manually or be done with prompts. So, it will maybe add some details to the image, which might make it better or worse. There’s more of this sense of play there too. It’s an endless process of getting the image perfect.
I think a lot of people don’t realise how much work goes into it either side of the prompt.
Well yeah, in theory it only takes two minutes to write a prompt. And you will get a result, but you won’t always like the result. It’s incredibly rare that I’ll be happy with an image that has come out of the blue. I have to have a creative direction for the whole idea in my mind. It’s not just robots – your own impact is the most important thing.
A lot of the AI generated images that we are used to seeing on social media aren’t really attributed to anyone or recognised as having an artist. You know because the Internet is so incredibly saturated with images. Images are often not explicitly attached to an artist, even though obviously they have to come from somewhere. Given this landscape, how do you think you’ve managed to build a reputation keeping your name attached to the work.
Why am I more popular than other artists? Well actually I do analytics a lot. I analyse everything that I do, and I always analyse the engagement that I get from people, from followers and otherwise. Whenever I post something, I always analyse like, Ok – this one people love, but this one people didn’t respond to. If there’s lot of replies to an image, then I can start to work out which content is good to go. I can post this kind of content more often rather than that one. And also, I know a lot of AI artists that do really amazing jobs and I love their images, but I know that they don’t have as many followers and engagement or popularity as my account. It’s because they tend to post all the images they personally love, whereas they don’t consider what might go viral on social media. I think my background in marketing and branding strategy plays a big role. Having done those sorts of jobs for maybe seven years means that I just know social media really well.
Well, on that note, you have worked with a variety of different brands, including several luxury fashion houses – Gucci, Versace, Valentino, I could go on. How do these collaborations tend to come about? Do the brands approach you or vice versa?
I never do sales. I don’t like to sell myself, so I avoid it as much as possible. Basically, they approach me, they find me somewhere. The people from Gucci told me, “OK, we know you are going to collaborate with other fashion houses.” And I was like, how do you know this? And they told me: “Your name spreads really quickly and we talk about you with each other.” So, I understood, OK. They talk to each other about AI artists, social media influencers and so on. They pass each other good people to work with and that helped me a lot to collaborate with a lot of fashion brands.
That makes sense. And are there any fashion brands that you’d perhaps want to avoid working with?
To be honest so far, I have loved the creative process with all of the brands that I have worked with. But maybe there are some fashion brands that I wouldn’t want to work with at all. My dream is to get to work with Balenciaga. They very much suit my approach, my style. I love them. Yeah. I think this could actually happen soon.
Yes, that definitely sounds like a good fit. One thing I was wondering about when working with these big brands would be issues of ownership and intellectual property. Can the legalities of copyright, image rights and such become a bit murky with the introduction of AI into the creative process?
With AI, no, they actually have a really strict legal system. Throughout the process with Gucci whenever I had created something. And the whole team had approved the image it had gone through several levels of approval process, it would end up being the legal team that would look at the images and say, “No, we cannot approve this. We have some minor problems here.” I don’t know. And then I’d have to start all over again. Yeah, but in general in Midjourney you have lots of packages that you can purchase and there is this business package that you can purchase, and the images are all yours and the copyright is yours. They are not public; they are in stealth mode. No one can see what you are prompting. No one can see the output of your images. This package is really good for working and the clients usually ask me if I have this business package from Midjourney where the images are yours. This is also the case with Photoshop Beta. Do you know it? It’s a tool where you can basically add anything to an existing image just by prompting it. Let’s say I have an image of a cat with a hat on and I want to change the style of the hat. I just select the hat and type, I don’t know, green furry hat, and then it creates a green furry hat. But again, there can be copyright issues here so you just have to have the business plan to make sure you’ve got all the rights to your own images.
At what point does the idea become intellectual property? Would you have ownership of the actual prompt itself?
Yes, actually. Even when the idea is small and not public, everything belongs to you and that includes the prompt. I could even sell all of my prompts if I wanted to, but I don’t.
We have established that your advertising background has definitely informed the way you operate as an artist. Would you see your creative output now as a continuation of your previous career path or a departure from it?
I quit all the office jobs that I worked here in Armenia. After advertising I had been at a huge IT corporation that had a big focus on AI. We were creating chat bots, so I was very up to date with the developments in the world of artificial intelligence. I was one of the first people who knew that GPT-3 was being released. I think being on the inside helped me to get into this world a lot faster. After really digging into it I realised that I understand this sphere and I quit my office job after about six months. I just realised that I couldn’t really combine the two lines of work. I’d be going into the office and doing all my work, like my personal work, and I knew I’d have to put one of them aside so I decided to focus on the art.
A lot of people involved with the art world have some serious concerns about AI and the consequences that it could have on the livelihoods of artists. Although it clearly hasn’t threatened you, what would be your response to fear and apprehension that surrounds this relatively new medium?
Well, actually, I’m an artist in the more traditional sense myself. As an illustrator I think that it’s the opposite. As long as you know how to introduce yourself, and as long as you know how to present yourself as a really cool artist on social media or wherever then AI art is just like any other artform. You just have to stay consistent in your work and be prepared to adapt yourself to new tools. As a graphic designer and an illustrator, AI is just another new tool than I can use to add to my existing skillset. I have work commissioned as an illustrator without AI and I don’t feel at all threatened. If anything as time goes by, the illustrative arts actually created with human hands are only going to become more appreciated. Nobody wants to lose them, and AI doesn’t mean that we have to.
So would you say that AI is just another tool like the pencil or the paintbrush?
It’s a tool, a very powerful tool. I remember when digital art was becoming very popular. I bought myself a tablet when I was a teenager and started to make digital works and I remember the reactions everyone had. People insisted that it wasn’t real art and that art had to be made with canvas and oil or pencil and paper. And I always just thought there was no real difference. What is the difference? You can see both of them. This same reaction happens every time a new form comes in. These days I get the same reaction with AI in comparison to my own digital art. Because of this cycle I know that in a few years AI art will be much more acceptable and respected. Life goes on and time goes on, you just have to adapt with everything that comes along.
That reminds me of when David Hockney had a big show at the Royal Academy back in 2012 and in these spaces where there would normally be huge oil paintings instead they installed loads of screens showing his iPad drawings of landscapes. It was really revolutionary at the time but now it wouldn’t seem shocking at all.
Yeah, I’m sure that he got some backlash too, some negative reactions. It’s normal people are afraid of something new. Actually, I know this by my experience. I remember my friends were telling me whenever I was discovering this AI image generators. I was going to do some sketches of plans for my colleague who works as a 3D artist. I was going to create some sketches for him and then he was going to transform them into 3D models. It just so happened that that day I had discovered Dall·E so I used it to create these really beautiful sketches. But when I sent them to him, he was annoyed that I had only spent a day on them, and he couldn’t understand how it had been so quick because obviously sketches take time. When I told him that I had made them using AI he told me to go away and that he couldn’t use them because they weren’t real art. He said he didn’t want to mix AI into his project. I just told him he could do whatever he wanted and I could do what I wanted. Then within six months he had changed his mind completely and now he actually appreciates my art very much. We laugh about it now.
Glad to hear it. I think the quicker the technology changes, the quicker people’s reactions change too.
Of course. At first, he was afraid of these tools because he thought that he would lose his job. He thought that as a 3D artist he wouldn’t have any clients anymore because they could just do it themselves with AI. But it’s not that simple. The client can’t just do it themselves. These are tools with skills that need to be learnt. Clients want to hire somebody who knows AI, and he realised this very quickly.
Yeah, I’ve tried to make some AI art a few times and it’s absolutely terrible. I’ll take two random pop culture phenomena and try and mash them together into a prompt, but they generally end up looking pretty horrendous.
Well, this is a good example. Actually, the client understands that it will take a very long time for them to achieve anything noteworthy and that is why they decide to hire someone who understands the technology better.
Moving on from the technology to the actual artwork itself, what genre would you consider your artwork to fall under? Would you call yourself a surrealist?
Well, the images are definitely very surreal. The ideas themselves are surrealistic and I love surrealism in art, but in terms of the image quality I really always try to make them as realistic as possible. This is in effect what I’m trying to do, surreal ideas but with hyper realistic image quality.
And would you say that you have any major influences in terms of specific artists or movements? Or do you just try and be as original as possible?
Well, I never use any specific artists or artworks for references or anything like that, only contemporary art influencers or old Asian artists who died a long time ago and they were all painters. My number one influence would have to be Salvador Dalí. Other than that, I love films. Specific camera techniques are really influencing me a lot in terms of image quality, different types of framing, the angles. When I first discovered these AI generators, I started to learn about the angles and different kinds of scenes that film directors were using, and I was going and searching for why, for example, Tarantino uses these low angles. This is how he achieves a particular effect and so on. This is how he creates these beautiful scenes.
So, do you work a lot of cinematographic terminology into your prompts?
Of course you can. About a year ago you could put all film terminology in terms of camera positioning and such into a prompt. But these days the technology is so advanced that you don’t really need to. In theory you could put everything into one sentence. But it’s difficult for you as the artist, rather than the technology, to be able to fit everything into one sentence.
Well, this question of fitting everything in brings me onto the idea of worldbuilding. Much like Dalí, a lot of your works feel like they expand beyond the borders of the image. It’s as if we are just catching a glimpse into an entire world that has been created. In your mind is there any lore, as such, or imagined worlds existing beyond these borders?
This brings me back to ideas taking root in childhood. I love cartoons. I love everything about fantasy and colour, cartoons and everything that is not real. And in my mind, I have many, many worlds. The wholesome characters in my daydreams and whole imagined worlds in my head spill out onto the page. I wanted to give life to these worlds and that’s how I started to draw and illustrate, and these AI works are just a continuation of that.
The pace at which AI is developing and changing art is showing absolutely no signs of slowing any time soon. Where are you heading in terms of developing and adapting on your journey with the technology?
Right. Sometimes whenever I’m thinking about this, it scares me a little bit. But usually, I have to keep this ethos that I’m going to adapt to anything and everything, wherever it may go. I really cannot imagine where it will go because people are already creating films with AI. It’s not that advanced at the moment, but I know in a few years we are going to be seeing entire feature length films created entirely with AI. As well as films, there will be upgrades in the overall working process in all spheres: music, design, art. Also, in things like project management, ChatGPT is releasing a new feature that has a personal assistant who you can train to pass certain tests and perform specific tasks. This is something that could massively help people in their creative lives, in work, and maybe even their personal lives too.
It certainly looks set to transform the world. And if these are the sort of projections we’re getting for the next couple of years, do we have any idea of what the AI possibilities would be in, say, ten years?
People are saying that AI will, I don’t know, control us. It’s only going to get smarter and smarter every year. And yes, I’m sure it’s going to transform into this sort of super brain that knows everything in the world. Who knows everything in the world? Who could? Imagine you have all the data from the Internet inside your brain – you are going to be a genius. What I’m saying is how do we know that AI doesn’t already control us now? AI can seriously manipulate us through social medias without us even realising. But the thing is there’s always someone behind it who has to make the commands for the AI to perform this task or that. It’s always the human who has to push the button and make a command for AI. But in ten years, maybe at that point AI will have reached a level of intelligence where it can just make the commands for itself. A lot of these ideas sound very futuristic and unrealistic. But everything is possible. For now, who knows.
Such turbulent times. Any final words?
All I have to say is that I wish people weren’t so closed minded or scared of these prospects. It pays to be open minded. People should try it because it will seriously help them a lot in life.
