Goodbye, Midjourney | Five Practices to Help Students Navigate AI Consciously

Goodbye, Midjourney | Five Practices to Help Students Navigate AI Consciously
Photo: Dave Edwards

I've struggled with how to create images for the past couple of years. Without the budget for a staff designer, we started experimenting with Midjourney to create abstract images. We initially tried creating images that conceptually linked to article content—but, over time, these images felt too fake and uninspired. And they felt like the kind of images we would much rather have a human create if we had the budget. More recently, we've changed to abstract biological patterns in an attempt to distance ourselves from the human conflict. But that hasn't entirely worked and it has surfaced a more personal problem.

Through our research on human-AI adaptation, we've identified three psychological traits that shape how people relate to AI systems. One of these—identity coupling—describes how closely someone's sense of self becomes entangled with AI interaction. When identity coupling runs high, AI systems can feel like genuine creative partners that enhance your capabilities. But they can also blur the boundaries between your ideas and artificial output until you lose touch with your own creative voice.

After a couple of years of working with Midjourney, I realize that the tool's design has trapped me in problematic identity coupling. Yes, I am solely responsible for the prompt, but the resulting image isn't something that satisfies my creative authorship. I toss an idea into the ether and then lose all creative direction and control. The system responds with something I couldn't have made myself—sometimes impressive—but it's never quite what I envisioned, and never feels like it emerged from my creative process.

This isn't about the quality of Midjourney's output (although that's also an issue). It's about the psychological relationship the tool creates. No matter how carefully I craft prompts, I'm outsourcing the essential creative act—the translation of vision into visual form. My identity as a creative person becomes so coupled with Midjourney's processing that I can't feel creative without it.

So we're changing our images again—to our own photographs. We won't attempt to match photographs to article content but, rather, share something that's in the world around us. Our articles are personal expressions of our journey studying AI and our images will be personal expressions of our journey in the world. While we won't present you with professional-quality photographs, we will present you with photographs that express our authentic creativity.

I have no issue with creative technologies. My most formative career moments relate to my role in creating tools like Final Cut Pro, GarageBand, and Keynote. These tools digitized processes that some felt took them away from their form of creativity while millions of people found their creativity through and with these tools.

The crucial difference is creative agency. Traditional creative software extends your capabilities while preserving your creative process. When I edit in Final Cut Pro, every cut reflects my judgment. When I compose in GarageBand, every note choice is mine. The software handles technical execution, but I remain the author of my creative decisions. With generative AI tools like Midjourney, there's a black box moment where your input transforms in ways you can't predict or guide. Your creative intention enters the system, but what emerges isn't yours. Human creativity is a beautiful black box full of wondrous mystery (how did Chris Stapleton write that song?!?). But, I believe the creator needs to be in the black box to be the creator. So far, generative AI tools are designed to keep me out of the black box and away from authentic creativity.

That said, I believe it is possible for generative AI tools to enhance rather than replace human creativity. There's nothing preventing authentic creativity with generative technologies—and I've started designing how. But current approaches prioritize efficiency over creative partnership. Their creators have fallen in line with the Silicon Valley obsession with seamless, frictionless automation. What we need are new ideas driven by creativity and meaning-making, not just productivity and convenience.

Until AI tools offer that kind of conscious creative partnership, we'll enjoy photographing what captures our attention—and hope these images speak to some of you too. They may not be the same quality as a true pro (or even, perhaps, an AI) but they will be ours, in a way that matters. They will represent moments we noticed and wanted to share. These photographs will hopefully provide flickers of human connection rather machine mediations. Something less efficient, but more meaningful.


Five Practices to Help Students Navigate AI Consciously

Educators are reeling. AI is disrupting education as much as any other industry, and everyone is scrambling for answers. Recently, the Atlantic published an article by a former Bates College professor calling for not only banning AI but essentially banning all technology for students.

My first reaction: can you imagine telling students they can't have a smartphone on campus at all? Have you spent time with students recently? (Note: our children would likely say the same thing about us!)

We are cautious about AI use in education and its effect on learning. But we believe it's better to educate students about how AI affects their learning rather than prevent them from having the experience. AI will be part of these students' futures—and learning how to learn with AI (or when to learn without AI) feels crucial.

Helen's piece on this topic is brief but important. As she describes, students unconsciously assign AI different roles—from idea sparker to complete outsourcer. When they default to outsourcing developmental tasks, they bypass the intellectual struggle that builds real capability, creating false confidence that grades don't capture.

We offer five practical strategies for professors: having students name AI’s role before assignments, checking when confidence depends on AI performance, tracking genuine cognitive contribution, designating AI-free skill-building zones, and preparing students for hybrid intelligence futures.

It's less than 6 weeks until...

The Artificiality Summit 2025!

Join us to imagine a meaningful life with synthetic intelligence—for me, we, and us. In this time of mass confusion, over/under hype, and polarizing optimism/pessimism, the Artificiality Summit will be a place to gather, consider, dream, and design a pro-human future.

And don't just join us. Join our spectacular line-up of speakers, catalysts, performers, and firebrands: Blaise Agüera y Arcas (Google), Benjamin Bratton (UCSD, Antikythera/Berggruen), Adam Cutler (IBM), Alan Eyzaguirre (Mari-OS), Jonathan Feinstein (Yale University), Jenna Fizel (IDEO), John C. Havens (IEEE), Jamer Hunt (Parsons School of Design), Maggie Jackson (author), Michael Levin (Tufts University, remote), Josh Lovejoy (Amazon), Sir Geoff Mulgan (University College London), John Pasmore (Latimer.ai), Ellie Pavlick (Brown University & Google Deepmind), Tess Posner (AI4ALL), Tobias Rees (limn), Beth Rudden (Bast AI), Eric Schwitzgebel (University of California at Riverside), and Aekta Shah (Salesforce).

Space is limited—so don't delay!

Learn more

Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to Artificiality Institute.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.