🎨AI Art: Perspectives on  Artistry and Creation🖌️
21:05 Share

Welcome to a new special deep dive from AI Unraveled. This is the show created by Etienne Newman. He's a senior engineer and also a passionate soccer dad up in Canada. If you're finding these explorations valuable, maybe take a quick moment to like and subscribe on Apple Podcasts. It, uh...

It really does help more people find us. Oh, and definitely check out the show notes. We've got a referral link and a discount code in there. You can use it to sign up for Google Workspace, get your hands on some really powerful AI tools like Gemini PRO, notebook loans, Teams, that kind of thing. Also, if you're, you know,

You know, aiming to nail those tough professional certifications, you should look into Etienne's AI-powered JamGadTech app. Seriously, it's designed to help you master like up to 50 plus performance-based questions and those tricky simulation-driven certs that are really in demand now.

So today we're diving into, well, the pretty captivating and frankly fast-changing world of AI artistry and creation. We're pulling from a bunch of articles, discussions, different sources really, that offer quite diverse views on what's clearly a big moment in cultural history. Exactly. And the core question today, it feels really fundamental, is

In this age of AI, what does it actually mean to be an artist? And, you know, what even is creation now? Our sources get into the words we use, like, what do you call someone using AI for art? Yeah. And the nuts and bolts of the creative process itself, that whole interplay between having an idea and actually making it happen. And whether, you know, asking AI for art is anything like working with a human artist. Yeah. And we'll also dig into the legal side of things.

uh, especially how copyright and authorship are being re-evaluated because of AI-generated stuff, and of course the bigger philosophical questions it raises. Things about, well, creativity itself, the role of intention, even what makes humans unique if AI can do these things. Plus we've looked into online communities, places like Reddit,

to see how people are actually arguing about this stuff right now. Right. So our mission here is to really unpack these debates, maybe find some surprising angles and just map out the different perspectives on AI art. We want to give you a clear perspective

hopefully engaging understanding of this whole pivotal moment, but without making your head spin. Okay, so let's get into it. The first sort of sticky point that comes up again and again is the term AI artist. Can someone using these tools legitimately be called an artist? What do the sources say?

Well, it's definitely a hot topic. And yeah, the sources show strong arguments both ways. People who are okay with AI artists tend to stress that, look, the human imagination, the vision, the intent, that's still crucial for directing the AI. They see the AI as just a very advanced tool, you know.

like a camera or maybe a synthesizer. The creative spark is human. Right. It reminds me of past debates, photography, digital tools. The tools change, but the person using them directs things. And supporters also point out there's a whole new skill set emerging here, right? Like prompt engineering, knowing how to choose the best outputs, refining them. It's a new kind of artistic practice, requires learning, sensibility. Exactly. But then you have the counter arguments, which are also pretty compelling. Critics often say, well, hold on. The

The AI is doing the actual artistic labor, isn't it? It's rendering the image. The human gives a prompt, sure, but the AI translates that based on its training. And there's the whole hand skill argument, you know, the years someone spends mastering painting or sculpting. For a lot of people, that physical craft is, well,

fundamental to being an artist. Yeah, you could totally see that point of view, especially from artists who've put in those years. Yeah. And then there's the originality question. It's a big one. Yeah. Because the AI models learned by basically mashing up tons of existing human art. Can the output really be called original? And this whole naming fight, it really boils down to different definitions of art and skill, doesn't it? If art is mainly about the final look and the intent behind it, maybe AI artist works.

But if art is fundamentally tied to the physical human process, mastering a medium, then it's trickier. So it's not just a simple yes or no. Probably not. Maybe it's more like a spectrum of creativity, acknowledging the human input, but also what the AI brings, which is different. A spectrum. I like that framing. Yeah. Okay. So thinking about that process, the idea versus the making of it, it often gets simplified. Human has idea, AI makes it. But the sources suggest it's...

Well, messier than that. Definitely messier. I mean, yes, the human starts with the prompt. But bodies like the U.S. Copyright Office have pointed out that the prompt itself often doesn't really dictate the final expressive details in the AI image, which shows how much the AI's own processing shapes the result. Oh, right. You might have an idea, but how the AI interprets it through its sort of black box process really influences the outcome. And that can impact how much you feel you authored it.

These models, they're trained on just massive amounts of image data, right? They learn statistical links between words and visual features. So when you prompt it, it's just generating what's statistically likely based on those patterns. Sometimes that leads to, well, surprises. So it's not always a straight line from your brain to the screen. But then the flip side is AI can also be a tool for generating ideas, can't it?

Artists using it to quickly visualize concepts, explore styles they wouldn't normally think of. Exactly. It can be like a brainstorming partner. You can prototype ideas super fast, explore all these different avenues. And that leads to this sort of back and forth, this iterative prompting loop that

The AI's output isn't just the end goal. It can actually feedback and change your original idea. Oh, interesting. You generate something, see an element you like, maybe something unexpected, and then you tweak your prompt based on what the AI gave you. It becomes this kind of dialogue where the idea and the execution get all tangled up. So the machine influences the human too. It's not just one-way traffic.

And this idea of collaborative emergence, where the AI's unpredictability is actually seen as a good thing. Right. The artist shifts roles slightly, becoming more of a curator, maybe selecting, responding to the machine's unexpected gifts, you could say. Embracing the randomness a bit. Yeah. Letting go of total control. And if you think about conceptual art, where the idea is the main thing, you could argue the human prompter, the one with the core concept, is still the primary artist, even if the AI helps shape the final look.

the focus shifts from the hand to the mind. That makes a lot of sense. Okay, another angle the sources explore is comparing AI generation to commissioning a human artist. Superficially, it's similar. You describe something, you get art back.

But the differences seem pretty huge. Oh, absolutely. Fundamental differences. When you commission a person, you're tapping into their unique viewpoint, their style they've developed, their specific skills, their judgment honed over years. It's usually a collaboration, a dialogue. They bring their own life experience to it. Whereas with an AI generator. Yeah.

You're interacting with software, right? It's working off patterns and it's training data. It can make amazing things, sure, but it doesn't have independent artistic judgment or consciousness or personal history like a human does. The dialogue is more about tweaking parameters than chatting with a creative partner. Exactly. And the sources lay out these differences pretty clearly across a few areas. Things like creative control, the type of skill involved, how iteration works, cost, time, originality.

originality, the relationship itself, emotional connection, how predictable it is, and definitely the ethical side. Tell me more about those contrasts, like iteration. Well, with AI, iteration can be incredibly fast. You can generate dozens, hundreds of variations in minutes. With a human artist, it's more deliberate sketches, feedback rounds, revisions. It takes time.

And cost, obviously. That's a big one. Huge difference. AI is generally much cheaper, maybe a subscription fee. Commissioning a human reflects their time, skill, materials, reputation, usually a much bigger investment. And originality in style. A human artist brings their unique, evolving voice.

AI output, while sometimes surprising, is always derived from its training data patterns. It doesn't develop a personal style in the same way. Right. It learns differently. Fundamentally differently. The relationship is software interaction versus human collaboration. That affects the emotional depth too.

A human can connect with your intent, infuse genuine emotion. AI can't replicate that. What about predictability? AI can be quite unpredictable, actually, especially with simple prompts. It takes skill to steer it. With a human artist, once you know their style and communicate well, the outcome is usually more predictable.

And ethics, that's a whole minefield with AI from training data to job impacts. Very different from ensuring fair pay for a human artist. For sure. And it's worth saying neither process is monolithic, right? There's variation in both. But the core difference is that creative agency, that human capacity for learning and artistic growth. Artists evolve. AI models get updated by developers. It's not the same kind of growth. So it really depends what you're looking for.

that unique human collaboration, or maybe the speed and directness of AI realizing a vision. Okay, so given this whole debate about who's an artist, it makes sense people are looking for other names. Our sources mention a listener suggestion. Art prompt engineers, or APEs.

What about that and other terms floating around? Yeah, the search for alternatives is definitely happening. Art prompt engineers, that term emphasizes the technical skill, the precision needed for good prompts. Engineers suggest, you know, a systematic approach, understanding the AI's mechanics. Right. Other terms pop up too. AI art director focusing on the vision and orchestration part.

prompter or prompt artist, very direct about the input method, or things like AI-assisted artist or just digital artists using AI, which kind of hedge their bets, acknowledging the tool without making grand claims.

The connotations are interesting. Engineer sounds technical. Director sounds visionary. But prompter. That could sound a bit basic, couldn't it? Like you're just typing words. It can, yeah. Some people apparently find terms like prompt artist a bit dismissive. Like it undersells the creative input. So choosing a title isn't just descriptive. It's almost...

In a way, yes. Using terms from established fields like engineering or directing can be a move to legitimize the practice. While simpler terms might be used to differentiate it, maybe even downplay its artistic status, it really reflects those underlying disagreements about skill and value we talked about. So no easy answer. The best term might just depend on how someone is using the AI, right? Simple prompts versus complex workflows involving lots of refinement.

That seems likely. As these practices get more varied, we'll probably see more specific language emerge to describe different roles and levels of engagement. Now try and get a feel for the public mood on this. The sources looked at online discussions, Reddit especially. What did those reveal? Yeah, those forums are definitely where the arguments are happening live. There was one thread on Ray Wars titled something like,

AI artist doesn't make sense to me. And that really captured a lot of the skepticism. People there often separated the human idea from the AI's execution. They kept comparing it to commissioning an artist, basically seeing the AI user more as a client giving instructions, not the artist themselves.

Questioning the control, the real input. OK, so skepticism there. What about the Ghibli example? Ah, the Ghibli discussion about that AI Ghibli art competition. That was intense. Overwhelmingly negative reaction. People felt it was deeply disrespectful to Studio Ghibli's whole ethos, the hand-drawn craft, the personal vision. It wasn't just about the images. It felt like a devaluation of human skill, even a threat to cultural heritage they cherish.

Yeah, that Ghibli thing hit a nerve. It tapped into something deeper about valuing that human touch, that unique vision from someone like Miyazaki. Seeing AI try to replicate it felt wrong to many people. And across both those discussions, you see recurring themes. Definitely some gatekeeping about what counts as art, who gets to be an artist. Absolutely. And a real tangible fear about human skills being devalued economically and culturally. Yeah.

Plus, those definitional battles trying to fit this new tech into old boxes. And often bubbling underneath are those ethical worries, especially about the training data. It's clear these online spaces are where the meaning of AI art is really being hammered out, reflecting genuine anxieties. Though it's probably fair to say not everyone in those threads fully understands how the tech works, which can add fuel to the fire sometimes. That's a good point. More understanding of the tech could add nuance.

Which brings us neatly to the really big philosophical questions AI art throws up. Like, can a machine actually be creative? Does it get beauty? Does it have artistic intent? Yeah, the deep stuff. Intentionality seems key here based on the sources. Traditionally, we think of art expressing human intent, ideas, feelings, wanting to provoke a reaction. Can an AI without subjective experience

really intend anything in that way. That's exactly the challenge thinkers like Amy Hsu point to. AI art makes us question our old definitions of authorship, originality, intent. If an AI image makes you feel something, is that the AI's intention? Or are you just projecting onto it? Right. And that leads to bigger questions about what makes humans special if machines can do things we thought were uniquely ours, like making art. It feels like AI is forcing us to look harder at ourselves, at what creativity even is.

Tobias Ries' idea of it being a philosophical rupture is pretty powerful, not just a new tool, but something forcing a fundamental rethink, maybe beyond just a human-centric view. Exactly. It pushes us to reconsider intelligence, agency, creativity itself, maybe beyond our usual human-focused lens. Interestingly, some argue AI art could even be philosophically good for us. There's this idea of the metacrisis and AI potentially boosting right hemisphere thinking, more holistic, intuitive stuff that's

maybe restoring some balance in our overly analytical world. That's counterintuitive, logical tech boosting intuition. Interesting thought. But

But then there's the worry, the flip side. Will our own creative muscles atrophy if we just let AI do it all? That fear is definitely there. Joanna Masiuska's quote about wanting AI for chores, not for art and writing, really captures that anxiety, the fear that it devalues things that give life meaning. And maybe even devalues personhood if AI becomes a creative peer. But again, maybe AI is that mirror.

By challenging our ideas of creativity, it could actually lead us to understand human artistry better, like photography did back in the day. It didn't kill painting. It changed it. It made us think differently about it. So despite all these big questions, the human role is still vital. It's not just push-button art. There's real skill involved on the user side. Oh, absolutely crucial. The AI renders, yes, but the human guides, refines, selects.

Prompt engineering, for instance, is becoming a proper skill. It's not just random words. It takes practice. Understanding the AI models, knowing artistic terms.

Sources mention research showing beginners often lack the right vocabulary for specific styles. So prompt engineering, that's crafting the text instructions carefully. Exactly. It's about learning how to talk to the machine effectively to get the artistic result you want. It's a skill you develop. And it goes beyond just the first prompt, right? People use other techniques. Definitely. Things like image-to-image where you start with an existing picture and

And in painting, filling in or changing parts of an image or outpainting, extending the image canvas, these give you way more control to shape the AI's output. Right. Fine tuning it. And then there's curation. This is huge. AI can turn out so many images incredibly quickly. Someone mentioned the potential for 100 million pictures in an hour. Yeah. So the human job becomes sifting through all that.

Using their aesthetic judgment to pick the ones that actually work, that fit the vision, that selection is a creative act. Choosing the signal from the noise.

Makes sense. And it seems traditional art, knowledge, composition, color theory, art history actually helps make better AI art. It really does. That knowledge helps you craft better prompts, guide the AI more effectively, make better choices during refinement and curation. Those fundamentals still apply. There's still that debate, isn't there? Whether these AI related skills really compare to the deep craft learned over here as in traditional art. Yes, that tension remains.

Is it developing a skill from scratch or skillfully using the knowledge already baked into the AI's training? It's a fair question. But AI can lower the barrier for people to just try expressing themselves visually. Maybe people who didn't think they could before. For sure. It democratizes visual expression in a way, but it also seems to have a high skill ceiling for those who really dig into advanced techniques, hybrid workflows, combining AI with other tools. So it's a broad spectrum.

Right. And often the public debate focuses just on that entry level prompting, maybe missing the deeper engagement and skill that's possible. Okay. So we've talked about the process, the philosophy. Yeah. But we have to touch on the really thorny ethical and legal stuff.

This seems like a huge area of conflict. Oh, absolutely. It's complex and evolving fast. Ethically, the biggest concerns are probably the impact on human artists. Job displacement, skills potentially being devalued, the whole economic disruption. Yeah, you hear that a lot. And the controversy over training data is massive. Were the images used to train these models scraped without permission? Is it copyright infringement? Is it exploiting artists' labor unfairly? That feels like the core battleground right now.

It really is. Plus, worries about biases in that data leading to skewed or harmful outputs.

Issues of authenticity, potential for deep fakes or deception leading to calls for clear labeling of AI content. Even environmental concerns about the energy needed to train and run these huge models. And the moral rights of artists whose styles get mimicked. It's a lot. And the legal side seems just as complicated, especially around copyright. Hugely complicated. In the U.S., copyright law has this human authorship requirement.

The Copyright Office has basically said purely AI generated works where a human just typed a prompt generally aren't copyrightable by the human.

So just prompting isn't enough creative input for copyright? Generally, no, according to their current stance. But it gets nuanced with AI-assisted works. If a human significantly selects, arranges, or modifies the AI output, those human contributions might be copyrightable. The Zari of the Dawn comic case is a key example here. Ah, right, where the images themselves weren't copyrighted, but the arrangement and text were. Exactly. There are arguments, like from Edward Lee...

suggesting maybe a bare minimum creativity test could work for AI art authorship.

But right now it's messy. There are ongoing lawsuits, attempts at legislation like the Generative AI Copyright Disclosure Act asking for transparency about training data. And it probably varies country by country, too. Definitely. The core challenge is applying old legal ideas about authorship to works generated from learned patterns, not direct human creation in the traditional sense. And artists unions must be getting involved. Very much so. Unions like the WGA, SAG-AFTRA,

equity. They're actively working to protect members. Key demands often include consent before using work for training or making digital replicas, fair pay, protecting moral rights, transparency about AI use and job security measures. So they're pushing for controls and compensation. Right. That push for transparency in training data, clear labeling of AI content. It's all about trying to regain some control and accountability.

It's interesting though, legally AI art struggles for copyright. That's a fascinating tension, isn't it? It potentially challenges the whole role of copyright and the value we place on human authorship if the market treats AI work similarly anyway. Wow. Okay, so wrapping this up, it's clear AI art has seriously shaken things up. Old ideas about skill, creativity, authorship,

They're all being questioned. Definitely unsettled the paradigms. And those lines between the human idea and the AI's execution are getting really blurry, aren't they? Yeah, that iterative process, the collaborative emergence. Yeah. It's not a simple division of labor. And the ethical and legal knots around copyright and training data are incredibly complex and far from solved. It feels like AI is this potent mix, part collaborator, part disruptor.

Which likely means the future isn't one single thing, probably a diverse creative landscape. AI might make visual expression more accessible for many while also becoming a super sophisticated tool for professionals who learn to master it. The future isn't written yet then. It depends on ethical guidelines, how the law adapts, public discussion. And the role of artists and their unions in actively shaping that future will be crucial. Maybe that uniquely human element, intention, emotion, lived experience,

will actually become more valued in art because of all this. It's a compelling thought, a counter reaction perhaps. So here's something for you, the listener, to think about. How is all this challenging or expanding your definition of art and creativity? And what role do you think this kind of technology should play in the future of how humans express themselves? Definitely something to keep exploring on this constantly evolving canvas.