26 Comments

Hi. I just came upon your substack. The illustrations are impressive but I do not see any person credited for this artwork. Did you do them yourself? I am wondering if they are AI generated. This is an important question for me to have answered before I commit to a paid subscription. Thank you in advance.

Expand full comment

Hi Ruben. There's a mix. Some of the illustrations are my own (for example https://simonkjones.substack.com/p/39-expeditions-and-interrogations or https://simonkjones.substack.com/p/35-accusations-part-1), some are stock footage (https://simonkjones.substack.com/p/29-the-creature-part-7), some I designed in Metahuman (eg https://simonkjones.substack.com/p/32-the-creature-part-10) and the rest are MidJourney (for example https://simonkjones.substack.com/p/51-zealots-part-3). I've also used a mix, with AI generation as a starting point for further manual work.

Some more thoughts on using MidJourney for the illustrations here (and the inherent ethical quandaries): https://simonkjones.substack.com/p/ai-book-illustrations-follow-up

I'm very curious to know more about your thoughts on the issue. AI generated images are contentious, and I'm thinking that might be a blocker for you? I'd like to know more about your position on the topic.

Thanks!

Expand full comment

AI art is suddenly everywhere now. Maybe it was everywhere already and I hadn't noticed. I'm very disturbed by it. I hate machine-made art generally because it's so destructive to my sense of what it means to be human. (I think the recent essay by Erik Hoel says it much better than I am able to: https://erikhoel.substack.com/p/ai-art-isnt-art ) But also I hate this sort of fake art because I made my living for many years as a graphic artist/art director/illustrator. I went to art school and studied painting and printmaking, found pride in my work and pride in developing my abilities. It's starting to look like it was a big waste of time. The lords of tech have absconded with my little sliver of expertise and social value. I don't work in the commercial art field anymore thankfully, but I am busy illustrating my serial novel. Even with the time saving tricks I learned when I worked professionally, these illustrations take me a couple of days each to produce. (The one I plan to release on Monday, took me five full days.) It's kind of a gut punch to see pictures, vastly more complex than the ones I labor over, done almost instantly by an app. I see these on your site and many other substacks. I can't imagine how it must feel to be a young art student at this point in time. Or an Illustrator or designer deep into their career.... AI art is a very negative development. It's a travesty and I wonder what will human artists do in the future? I suppose confine themselves to mediums that aren't digital; paint and canvas and flesh and bone. It's more likely however, that we'll all rationalize this as some great new development and we won't even notice when nobody knows how to draw anymore.

Expand full comment

I understand your concerns. There's a parallel development in AI writing which worries me personally - though at the moment AI writing is abysmally dull (if technically 'correct'), it's presumably only a matter of time.

That said, I don't know if it's a bleak as you fear. I hope it isn't. I'm using AI art to add some colour to my stories, and I find it useful as a way to add some visual flair, but I'm under no illusions. It's evidently AI art, and in that regard it's similar to the stock images I was sometimes using previously - technically fine, but without real intent. Borrowed.

I find AI images technically fascinating, but artistically rather empty. At least, when they're just presented as-is. My house has paintings and illustrations from all sorts of independent artists, each chosen because the work spoke to me or my wife or son in a specific way. The deliberate choices of the artists contributed to us putting them on our walls. I can't imagine putting an AI image on the wall, even if from a purely aesthetic point of view I've seen a lot of visually stunning AI images.

Mike Bithell argues the point well on a recent ep of the Play, Watch, Listen podcast (towards the end of the episode), where he notes that 'intent' is the vital missing ingredient. I think AI generation will settle down into being a useful tool, with the actually *interesting* art that comes out of it being when a real artist takes some AI generated source material and then does something....else with it. Much like photobashing or using stock images or reference photos.

It's certainly hard to predict. Maybe some people WILL hang AI images on the walls of their houses. But I don't see it destroying or taking over from traditional artists. Humans, I think, intrinsically need a human behind the art they experience - if I read a book or watch a movie or read a comic, knowing that A Person (or people) is behind it adds to the experience. Even if an AI could make a book or comic or movie as good in every way, it wouldn't quite be the same.

Or, rather, if it was able to do that, then AI would have reached some new level of actual sentience. At which point the debate shifts again, because 'intent' comes back into it. But I can't see that happening in my lifetime.

Even in cases like concept art, where you can imagine directors and producers leaning into AI art, it's still not going to be as good as hiring a proper concept artist. Sure, a director might use an AI generator to create a few quick ideas to help illustrate their thinking, but that would be followed up by the concept artist taking it (much) further.

All that said, I get how it's an unsettling time. My use of MidJourney certainly isn't intended to make anyone feel uncomfortable, and your comment has made me rethink some of my use of it. Specifically, going forward I'm going to try to find time to inject more of my own illustration into and on top of the Ai generated image. Thanks for that prompt.

Expand full comment

Well, I don’t like being the gloom and doom guy, but I give it about seven years. I think that within that time frame most if not all commercial artists will be replaced by AI. As far as you or anyone else using these apps to illustrate your work, I can hardly be angry. It’s a natural extension of what you do, it’s a need, and it’s expensive to hire a human. I certainly couldn’t afford a pesky human! I’m lucky that I can do the pictures myself. People make fun of the old Luddites. The term has a pejorative connotation, but today more than ever I sympathize with their struggle and I also understand it’s utter futility in the context of a capitalist economy.

Expand full comment

The real enemy is "good enough". We have become a society where we're we okay with quality being "good enough." I remember a long article talking about how mp3s are technically a terrible file format, with lots of compression loss, but the small size, and ease of sharing trumped fidelity (back in the Napster days). There were tons of other formats that were truly good, but none could get a foothold because mp3s were "good enough".

In the same way, while template video projects, and phone app auto-edit videos are inferior to what I can do, they are good enough for most no-budget small businesses. The same way stock photography isn't EXACTLY what you want for a project, but it's close enough. And soon AI art will be good enough to replace a good chunk of commercial artists, it's true, assuming they don't pivot and try to take advantage of it rather than let it make them obsolete.

Right now, every image that comes out of AI has that "look". No matter how amazing it is, I can easily tell if something's been done by AI. Once it starts getting out in the world and used in actual projects by name-brand creators, it'll start looking samey-samey and it'll take another advance of the technology to eliminate that telltale uniformity.

Even though all this stuff is encroaching on 'real' artists, the reality is that 99% of the people using this stuff are still too lazy to do anything creative with them. They'll be able to do stuff for themselves, but if someone else asks them to make something? They are clueless. It's because of this laziness that my own career will be extended a few years beyond what it should be!

Expand full comment

I work in video production (I have my own company), and have been seeing technology slowly encroach on my industry with increasing speed. By technology, I mean computers or apps being able to instantly put simple (but increasingly more complex) videos with the touch of a button. Lots of hand-wringing in my industry, but I'm not one of them. We can't put this genie back in the bottle. Doing everything one can to slow this down or stop it is a futile waste of time. The one thing none of this stuff can do (yet) is provide the personal touch and unique collaborative results you get from one-on-one client relations. My value isn't the specific video I shoot or edit, but being able to accurately tell the story the client wants to tell. What I've lost forever are any mom-and-pop small clients. They can do that on their own with simple online tools or even on their phones. But I still have a robust clientele of higher-end companies. But even that is probably on its path to extinction. Nothing I can do about it except be invaluable to my customers. I don't know exactly how it will relate to the art world, but they'll adapt, too. They'll find their patrons.

As to people doing art itself? That's the one ridiculous fear I keep seeing in this debate, is that suddenly people will stop trying to do art. For (checks notes) 99% of the public, art/music/writing isn't something we get paid to do. We do it because we MUST do it. And we always will. If you're just doing art to make money, then yes, you will likely stop making art. But people will always have a passion for expressing themselves in creative ways. Some of those people will even get to make money at it. AI art will just be another way to make stuff.

Now here's a real-world personal situation that illustrates both the positive and negative side of this. A couple of months ago, I decided I was going to start writing my long-gestating fantasy story as a serialized novel. I have a bunch of kingdoms and such that all needed some interesting designs. I could just write it all out, but I'm a visual person and like having concrete images to draw inspiration from. So I started reaching out of places like Fiverr to see if there were any affordable artists who could take my basic descriptions of my armor, creatures, geography, etc. to see if they could give me some basic concept art. The prices that were coming in on even the low end were, for me, prohibitive. $50 for just a rough sketch of a helmet. $200 for a very rough digital painting of a creature. Etc. It would cost me thousands of dollars to concept design my world. If I was a pro writer and this was for a paid project where this sort of art was necessary, then sure. But I'm a nobody, writing a novel for my own enjoyment. So it SEEMED that having any sort of concept art for my world was a no-go. Then LITERALLY the next day, I discovered all the AI-art stuff, and suddenly, the art design of my world went nuclear-powered. 95% of it isn't anything I'd post as "art" for my story, especially since even with my best efforts (I consider myself an AI-art wizard now), it's nearly impossible to get the same thing (like the way armor looks) in two images in a row, or ever. But what it IS, magnificently, is concept art for scenes and locations and moods, and tiny details, and colors, and terrifying aspects of monsters, and on and on and on, of things my brain hadn't even imagined. It's just stunning some of the stuff it's come up with and some amazing trajectories its sending me on.

So the question is, is it wrong for AI art to be able to come up with the equivalent of tens of thousands of dollars of concept art for me, even though it's not money taken out of artists hands because my budget wasn't tens of thousands of dollars. It wasn't even tens of dollars. None of what I was given is what I wanted, despite spending countless hours tweaking my art prompts. But it's given me so much that FEELS like what I asked for and is perfect for my purposes, which is stuff I can use for writing inspiration. Is the best answer that I just shouldn't be able to have ANY art if I can't afford to pay the going rate? If I want to describe details about the Eifel Tower, do I need to pay a photographer the going rate to go snap a picture of it for me? Or is it ethically okay for me to just Google it and use any number of photos I don't own as my inspiration? Do I need to purchase a print of the Mona Lisa, or pay a royalty to be able to look at a picture of it in order to accurately describe it in my story?

So many angles to this debate. Here's a wrinkle. So it's exhausting and frustrating getting AI art even close to what I want it to. I'd probably be happy to pay a skilled AI Whisperer to create art for me that more closely matches. I wouldn't pay $200 for a landscape, but I'd pay $50.

Expand full comment

Agree with that! Also, I really enjoy talking to artists, or reading interviews with them. No matter how good AI gets in terms of the raw output, we're a long way off them having engaging personalities in that regard. Getting to know an artists' catalogue is immensely satisfying. Getting to know an AI generator's strengths is more like becoming familiar with a tool. Very different.

MidJourney has advanced massively just in the last few months, so this is going to be a fast moving landscape. I don't see it as quite so apocalyptic - but it's hard to predict.

I do think there is a mid-level of artists that are likely to be affected, just as you mention the 'good enough' scenarios in another comment. Given the strengths of something like MidJourney, stuff like the "ethereal painting of wolves howling at the moon" illustrations that you see on t-shirts in shops in Glastonbury (to pick a random example) if probably over.

Expand full comment

AI art is going to get even more complicated because it's becoming increasingly easier to upload and use your own photos to be the basis for generated art. More difficult in that it'll be much easier for AIrtists (can I copyright that term?) to say "I made this." Already, I feel I can claim ownership and artistic credit for AI art I've created, especially when I spend hours and hundreds of generations of tweaking to get it as close to what I want as I can. And this is before supplementing it with my own artistic skills in Photoshop. It takes skill now to create something in MidJourney that doesn't look like every other thing created in MidJourney.

Interesting discussion that will be ongoing for quite some time, I think.

Expand full comment

Yeah, and the technology is evolving so rapidly, as you note, that it makes the discussion difficult. The landscape is shifting beneath our feet.

In my use of MidJourney to illustrate this newsletter, I very much consider myself a loose curator, or maybe a researcher, rather than an artist. There's definitely some effort involved in trying to generate what I want, and it often doesn't work. The process feels more like searching for stock images, or researching through documents, rather than being creative.

That shifts of course if you then take the AI generated image and do further work on it. At that point, it becomes one of many potential inspirations. I do think I need to do more of that - work on carving out the time to create my own take with the MidJourney image as a reference.

Expand full comment

Well done. Thank you!

Expand full comment

Really interesting! Thanks so much for this.

Expand full comment

Illuminating! Thank you for doing this work.

Expand full comment

Thanks for this resource!

Expand full comment

Thank you for sharing this, Simon! This is a wonderful breakdown. I plan to do a similar exercise toward the end of this year for my subscribers, and this is a great starting point for me to figure out what questions to ask.

Expand full comment

Ah, I'll keep an eye out! It's so easy to work in isolated little bubbles, but since I moved to Substack I've discovered a really vibrant community of writers. Sharing tips helps everyone, I hope.

Expand full comment

Mine will definitely be more reader focused as opposed to how people are writing. I have about 40-50 who are not writers, and approximately 100 that write full time or as hobbyists. It should make for some interesting feedback.

Expand full comment

It’s so cool to see what other Substack fiction authors are doing, thank you for putting this together! (P.S. Have you thought about making that PDF a Substack post and pay gating? Super valuable!)

Expand full comment

You're always one step ahead, Elle. ;)

I did wonder if I should make the report only available for paid subscribers, but I think this time round that would be too restrictive. However, further analysis/follow-up might be something I paywall in some manner.

Expand full comment

Thanks, this was really interesting Simon. I’ll participate next time around.

Expand full comment

Fab! It's definitely too small a sample to be drawing too many hard-and-fast conclusions from this one, but I think it's interesting nonetheless and something to build upon! Thanks, Tom.

Expand full comment

I bet you’ll have substantially more participants next time ... the momentum is building

Expand full comment

Hey Simon, thank you for taking the time and effort to pull this survey together. It is a really useful snapshot of what is happening in the space. I really hope you continue to expand on it. A suggestion: provide insights on what the stats mean and what trends may emerge in the space.

Expand full comment

Thanks, Rand. I definitely hope to expand on it in the future.

Expand full comment

But this is the survey by the writers. This won't highlight what people are actually reading and what genres are in trend.

Expand full comment

Correct: there's a lot of analysis out there of the market from the reader end. My intention here was to focus very specifically on the writers and what we're up to.

Expand full comment