Testing AI Comic Creation: Drawing Short Bande Dessinées with Images 2 in Multiple Languages

When I first saw the headline about AI image generators being tested for comic strip creation, my mind went straight to the stacks of Sandman trades gathering dust on my shelf back in Portland—specifically, the well-worn Volume 1 I bought at Powell’s on Burnside during a rainy April in ’09. The idea that a machine could now attempt to render sequential art, even in a rudimentary form, feels like watching someone try to teach a toaster to paint a watercolor. Yet here we are, with tools like ChatGPT Images 2.0, its predecessor 1.5, and the intriguingly named Nano Banana 2, all being put through their paces to see if they can grasp the grammar of comics: panel flow, visual pacing, the silent beat between speech bubbles. It’s a technical exercise, sure, but it also brushes up against something deeper—how we define creativity when the brush is held by an algorithm rather than a hand that’s held a pen since childhood.

The source material didn’t specify a U.S. Location, but as someone who’s spent years covering the intersection of tech and culture from the Pacific Northwest, I can’t help but anchor this discussion in Seattle. Why? Because this city isn’t just home to Amazon’s AI labs and a thriving indie game scene—it’s also where Fantagraphics Books has been championing alternative comics since 1976 from its Georgetown storefront, just south of the stadiums. It’s where the Short Run Comix & Arts Festival fills Seattle Central College every fall with creators hand-stapling zines and screen-printing covers. When we talk about AI attempting to build comics, we’re not just talking about pixels and prompts; we’re talking about a medium that, in places like Seattle, is deeply tied to community, craft, and a certain DIY ethos that resists automation.

Let’s be clear: none of the tools mentioned in the source material are anywhere near replacing a human cartoonist. The tests described—generating basic BD (bandes dessinées) strips in multiple languages—reveal more about the models’ limitations than their potential. Image 1.5 struggles with consistent character rendering across panels; Images 2.0 shows improvement in handling simple dialogue bubbles but still falters on complex backgrounds or dynamic angles; Nano Banana 2, whatever its exact architecture, appears to prioritize speed over coherence, often producing panels that read like visual non sequiturs. These aren’t flaws to be patched in the next update; they’re symptomatic of how these systems process visual information. They don’t understand why a close-up on a character’s eyes matters after a line of dialogue—they only recognize statistical patterns in the training data. For a medium as nuanced as comics, where a single tilted eyebrow can convey volumes, that’s a fundamental gap.

This isn’t just theoretical for Seattle’s creative community. Consider the impact on local art schools. At the University of Washington’s School of Art + Art History + Design, professors in the Illustration program are already grappling with how to teach foundational drawing skills when students can generate a passable sketch in seconds. It’s not about banning the tools—it’s about ensuring students understand the *principles* behind what they’re generating. Similarly, at Seattle Public Library’s Central Branch, the comics and graphic novels section (one of the largest in the Pacific Northwest) sees steady demand not just for manga and superhero titles, but for works by local creators like David Lasky (Oregon Trail: Road to Destiny) and Ellen Forney (Marbles). If AI-generated comics ever flood digital platforms, it could further saturate an already crowded market, making it harder for emerging artists to gain visibility—not because their work is worse, but because discovery algorithms might favor the sheer volume of AI output.

Yet there’s also a counterpoint worth considering: accessibility. For someone with motor impairments that make holding a pen difficult, or a non-native English speaker struggling to visualize a story idea, these tools could serve as a sketching aid—a way to get rough concepts down before collaborating with a human artist. I’ve seen this dynamic play out in adaptive tech workshops at the Seattle Public Library’s LEAP (Library Equal Access Program) lab, where tools are tailored to individual needs. The key, as always, is framing AI not as a replacement but as a potential collaborator—one that requires careful guidance and clear boundaries. A local comic artist I know in Capitol Hill uses AI-generated texture references for background elements in her webcomic, but she draws every character by hand, insisting that the humanity in the line work is non-negotiable.

Given my background in tech-cultural analysis, if this trend impacts you in Seattle—whether you’re a working artist, an educator, or just someone who loves comics—here are three types of local professionals you’d wish to consult, not for hiring per se, but for understanding how to navigate this shift thoughtfully:

Comics Educators & Program Advisors: Look for instructors at places like Gage Academy of Art or the Seattle Central College Creative Academy who emphasize *critical* engagement with AI tools—those who teach students not just how to use generators, but how to deconstruct their outputs, identify biases in training data, and preserve manual skill as a form of artistic literacy. The best ones will have recent curriculum updates addressing generative AI and can point you to student work that blends analog and digital methods thoughtfully.
Independent Comics Publishers & Editors: Seek out editors at local presses like Fantagraphics, Short Run’s imprint, or even smaller zine collectives. They can offer insight into what makes a comic resonate beyond technical execution—narrative voice, cultural specificity, emotional authenticity—and how those qualities remain uniquely human. Ask about their submission process and whether they’re seeing AI-assisted work; their criteria for what stands out in a stack (physical or digital) is invaluable context.
Digital Rights & Creative Technology Advisors: Consider consulting with professionals at organizations like New Media Rights’ Seattle affiliate or tech attorneys at firms like Stokes Lawrence who specialize in IP law. They can help clarify the murky waters around training data copyright, ownership of AI-assisted works, and ethical disclosure practices—crucial knowledge if you’re integrating these tools into your workflow or concerned about how your own style might be used in models you didn’t consent to.

Ready to uncover trusted professionals? Browse our complete directory of top-rated experts in the Seattle area today.