29 Comments

I look forward to seeing more of your illustrations. I know it's a new level of vulnerability to post them, but I can appreciate the effort to improve both your writing and art. I started creating new story images for my posts, but it's slow going. I find I enjoy getting to completion more than I do the final outcome, so we'll see.

Expand full comment

Yep, and I'm aware that my own art isn't going to be as immediately appealing or impressive at the MidJourney stuff. That's the trade-off, for now at least. It'll be more satisfying to produce, though! And I think (or hope) that for readers, having less competent but more 'me' art in place of super snazzy but generic AI images will be more interesting.

Expand full comment

Eric Hoel is right about ChatGPT being banal. He's wrong, and, I have to assume, deliberately hyperbolic, about Bing being evil. Bing has been trained on too many bad sci-fi novels in which such conversations occur regularly. ChatGPT is passing exams because it has the answers written on its sleeves. It's a pastiche machine, and exams are an exercise in pastiche -- no one is looking for original thought, they are looking for regurgitation of the required texts and that is what ChatGPT does best.

ChatGPT does not think. It ingests and partially macerates existing thought and spits it out again on demand. The only thought in the system is in the input, not the process. Thus the banality.

The great danger I see in all this is that the public may not detect the banality of AI art. The creative industries, after all, have been working for some decades now to dull our senses and our sensibilities. The books it turns out are simple tedious repetition of the exact same emotional triggers which have now been timed down to the page.

That emptiness you rightly detect in AI art is there is so much of the human produced art as well, and it is not an emptiness born of lack of talent, but a deliberately cultivated emptiness engineered by an industry the had instrumented attention and commoditized it. (Orwell predicted this in 1984.)

Apocalyptic writing is, of course, a core part of instrumented and commoditized attention, and Hoel seems to have mastered the craft of it and has turned it into an attentive audience and doubtless a nice living. He is very far from being alone in this. The question is whether there remains any other way of finding an audience these days.

Ironically, of course, that's a somewhat apocalyptical thought in itself. An apocalypses of apocalypticism is perhaps the thing we should worry most about. If there is anything to the recent reports about a developing teenage mental health crisis, I think we see the results of our apocalyptic attention grabbing. There are minds too naïve to see through this nonsense, minds who are being told they are heading into a perfect storm of apocalyptic threats just as they are at their most vulnerable as they prepare to leave the nest and start independent lives. Want to cheer the kids up? Stop telling them that the world is going to end.

The danger is not what the climate may do to us, or what the robots may do to us, but what we may do to ourselves.

Expand full comment

So well said. Bravo! I work in the climate/sustainability arena -- and there the doom & gloom is equally intense, although also very real. Still, there are a lot of people doing incredible things to turn the needle back to a healthier and saner world, and that's what I prefer to focus on.

Expand full comment

Thank you for this. As an illustrator, its been very disheartening to see its use in its current form with no oversight. This really boosted my spirits!

Expand full comment

Ah, I'm pleased! Someone challenged me in the comments on a post last year, who was also an illustrator, and I didn't really have a good answer. For a beginner illustrator like myself, AI images were exciting (and still are, in some ways), but increasingly it seems like it's been gone about the wrong way. If there's an ethically sound version of AI image generation in the future (if that's even possible...) then I might go back to it, but for now it feels more comfortable to stay clear.

I only recently found your Substack but I love your style. It's immediately a good example of why humans are still vastly more interesting than AI for this kind of thing. :)

Expand full comment

I think it can definitely be done ethically, and i can see how it could be used as a tool in the creation process, but it will require a lot of oversight and artist protections that will take a while to come in. I hope we get there!

Also, thank you so much! Haha yes my art is 100% human created

Expand full comment

I hear you and respect your decision even if I'm not remotely there yet. I use AI as THE most incredible collaborator I've ever encountered. Real creative humans are still the best, but as I'm sure you well know, their availability, and willingness to focus exclusively on MY projects makes relying on them as collaborators less than sayisfying.

I've always been more of a "tools aren't the problem" person. I don't think they should be banned just because some people use them irresponsibly. Make it easy for responsible use. Make it harder for the idiots or bad actors.

Easier said than done, to be sure, but the AI genie isn't going back in the bottle, so we have to figure out how to tame it. Or ask it for more wishes.

Expand full comment

That's the flip side, good faith users can get a lot out of Chat AI. Using new technology as collaborator is a really healthy and smart way to go. Humans, bless 'em, won't always make us the centre of their universe. Now there's a workaround, great example. Another example: if I provide 4 pages of original text and ask Chat AI to summarize in 12 dot points, it takes a couple of minutes to generate. That's a real time saver, my time. It's a little like magic, with good grammar.

Expand full comment

Yeah, and if someone comes up with an AI image generator that either doesn't rely on scraping thousands of artists' work, or which compensates or addresses the issue in some mutually beneficial way, I'll happily go back to using it as a tool. I think as long as so many artists are objecting, I'm not going to be entirely comfortable using such a tool.

If it was just me and the AI, it'd be fine, but by design at the moment, using an AI image generator has wider ramifications on other artists.

It's a thorny one!

Expand full comment

I think images and art are in a separate category, with distinct moral considerations, as well as being a whole different discussion about the merits and value of human endeavor and creativity.

Expand full comment

Ha, exactly. And I think Ai image generation in particular has the potential to be a very strong tool, especially for people like me who love drawing but lack some of the skills needed to accomplish what I want to do.

The problem currently, I think, is that there are awkward ethical concerns at the root of how these systems work. Even if my personal use is entirely ethical, are the systems at their core still reliant on using the work of other artists, some of whom don't want their creations absorbed in such a way? If there can be a system that either avoids or addresses that concern, I'll happily use AI image generation again for inspiration and as a tool. I've no idea from my layman's position whether it's even possible to separate the two, though.

This post certainly isn't my final word on the subject. I suspect the next year or two will shift the debate massively in all sorts of unpredictable directions.

Expand full comment

You had me at the source code. Purchased!

Expand full comment

Thank you!

Expand full comment

Your premise for "No Adults Allowed" reminded me of the Ray Bradbury short story (13 pages) titled "The Veldt." Now, when I say reminded me, I do not mean they are similar, beyond the issues of AI and kids. He wrote it for the Saturday Evening Post in 1950. I think you'll get a kick out of it, if you haven't already read it. https://repositorio.ufsc.br/bitstream/handle/123456789/163728/The%20Veldt%20-%20Ray%20Bradbury.pdf

Expand full comment

Ah thanks for the tip, will check it out!

Expand full comment

Simon, want to commend you on your decision to quit gen art AI and focus on your own art. To your point, the images feel somehow empty. Soulless. The good ones that I've seen are those honed with great detail and effort by actual human artists, who use the AI as a tool, not as the creator.

Why do I care about this, personally? I'm the daughter and the mother of artists. My daughter is just 12 but already a professional artist. When I think about the world she is growing up in, I wonder and worry about how she will be able to get her work out into the world, make a living, and enjoy this craft. Art and creativity are the very bedrock of the human spirit. I refuse to relinquish that to an algorithm, and it's heartening to see how many of us feel this way.

Expand full comment

Hooray! Right decision.

Expand full comment

Love that this has reignited your passion for drawing! So cool!

Expand full comment

Lot to unpack here. I join those applauding you in returning to original art. I've played with multiple AI generators (and have played most with Stable Diffusion implementations where I can at least feed it one of my photos or drawings as an IMG to IMG starting point), but they either frustrate me (spend enough time refining a prompt where I could have just done it myself) or bore me (get a technically solid result that's just meh). I've generated hundreds of images, only to delete them the next day.

The AI will never understand the "proper" output to the prompt, "a white cat facing away from camera in a Snowstorm" is a white frame and a pink dot.

Also, it's a bit frustrating the current code is biased towards European influence. Feed it a source image of one of my white friends and I get something photoreal. Feed it an African, Asian, >other< source and the AI either turns them white or I get nightmare fuel - unless I add additional prompt language. I spent a whole week testing that. *(no more line breaks) *And plenty of Prompts where the AI just won't give me what I want. Feed any AI HP Lovecraft's description of Cthulhu and you'll get the octopus head - but no body, no wings. The AI is still pretty stupid. *Same with ChatGPT. I've run experiments, but, ultimately quickly hit the point where asking for elaboration on X means Y completely goes away. After a few iterations it's better to just write it my own damn self. I'm better. Period. And I'm only ok. *The tech is new. It will improve, but I suspect I'll always be a Luddite here. I just don't get personal satisfaction from the AI tools. *Don't think I got back to you over the weekend, but the articles, also linked here, were excellent reads. Thanks for sharing!

Expand full comment

Much like how it became clear over the last decade that The Tech Bros aren't really best placed to solve complicated social issues, I'm pretty sure it'll also become clear that they're not really best placed to wade into the art world, either.

Expand full comment

Amen to that. I live in Silicon Valley... feel my pain, feel mah pain! :)

Expand full comment

Is it even possible to have non-tech-related conversations there? :)

Expand full comment

Haha yes it is but you have to make it so! In a way it's great to be a woman (and a mom) here bc you can talk about all kinds of things non-tech without the excessive expectations.

Expand full comment

I enjoyed reading this post, and for the record, I applaud your decision to do your own illustrations, even if they aren’t at the skill level you’d like. As a reader, I can appreciate that you are developing your talent in this area. Raw and unpolished is okay. Truly.

Also, how did you insert the “quotes” to Erik’s posts? Were these cross-posts? I’ve not used that feature yet so I don’t know how to do it.

Expand full comment

Thanks!

The links to Erik's name were done by typing an asterisk @ followed by his name. The embedded links to the posts were simply by pasting in the Substack link - Substack automatically converts Substack links into those nice little stubs.

Expand full comment

Thanks for your help. I forgot about using the @ sign. Love how Substack automatically creates those nice little stubs. I'm definitely going to make use of this.

Expand full comment

I couldn't read your post in its entirety but I can add that I do believe AI was programmed as capable of plagiarism. Sorry limited time to finish reading.

Expand full comment

I hear what you're saying about the graphics generated being a little dull, if not icky. Just starting to explore in BlueWillow to see what I can conjure and it's frustrating/impossible to get what I hope for.

Expand full comment