I spent two hours trying to ‘fix’ a puppy picture with AI. It taught me a valuable lesson about technology and time.

It started with a simple, wholesome idea. I have this picture of my puppy—ears flopped over, looking ridiculously proud of a stick he found. It’s a great photo. I thought, “You know what would be fun? Let’s make this look like a still from a Pixar movie.” It seemed like a perfect five-minute task for an AI image generator. Two hours later, I was slumped over my keyboard, filled with a specific kind of modern despair. I had fallen deep into a loop of pure AI frustration.

What I thought would be a quick, delightful project turned into a maddening cycle of “almost.” The first result was pretty good, but the eyes were a little… off. Creepy, even. So I typed, “Make the eyes less uncanny, more expressive.” The AI complied, but now one of his ears was shaped like a croissant. “Okay,” I muttered, “fix the ear.” The ear was fixed, but the stick in his mouth had morphed into a weird, lumpy banana.

This is the rabbit hole of AI frustration, and it’s a place I think many of us are becoming familiar with. Each prompt was a negotiation. Each “fix” was a compromise that introduced a new, bizarre problem. I was so close to the perfect image, yet every single attempt was just flawed enough to keep me hooked, certain that the next tweak would be the one that finally worked. It felt less like a creative partnership and more like I was arguing with a genie who was a master of malicious compliance.

The Downward Spiral: An Exercise in AI Frustration

Why does this feel so uniquely infuriating? I think it’s because the technology is just good enough to make you believe your vision is possible. Unlike a traditional tool like Photoshop, where your limitations are your own skills, AI presents itself as a tool with near-infinite skill. The only limitation is your ability to describe what you want.

But there’s a catch. We’re communicating with a system that doesn’t actually understand context, aesthetics, or why a banana-stick is weird. It’s a hyper-advanced prediction engine, assembling pixels based on statistical patterns from its training data. As explained in this excellent overview by How-To Geek, these models aren’t “thinking” in a human sense. When you ask for an incremental change, the AI isn’t editing the image; it’s often generating a new one based on a slightly modified understanding of your prompt. This can lead to unpredictable, often bizarre, results.

The process ends up looking like this:
* You have a clear goal.
* The AI gets you 90% of the way there on the first try.
* You spend the next hour fighting the AI over that last 10%.
* You either give up in a huff or settle for a result that’s “good enough” but not what you originally wanted.

Is AI Frustration the New Social Media?

As I finally closed the laptop, defeated by the puppy-croissant-banana monster, I had a thought that truly bothered me: this felt just as bad as mindlessly scrolling social media. Both experiences start with a simple intention—to connect, to create, to be entertained—but can quickly devolve into a time sink that leaves you feeling drained and unproductive.

With social media, it’s the “infinite scroll” designed to keep you hooked on the next potential dopamine hit. With generative AI, it’s the “infinite tweak.” It’s a gamified loop where the prize—that perfect image, that perfect paragraph—always feels just one more prompt away. This constant cycle of near-success and failure can be incredibly draining, mirroring the same psychological traps that make other digital platforms so addictive. Research from institutions like the Mayo Clinic has long pointed out how social media can impact our well-being, and it’s worth considering if this new creative loop poses similar risks.

I’m not saying AI is bad. It’s an incredible tool that can do some truly amazing things. But my little puppy project was a reminder that it’s just that—a tool. It’s not a magic wand. And just like any other powerful digital technology, our relationship with it can become frustrating and unhealthy if we’re not mindful.

For now, I’m going back to the original photo of my puppy. It doesn’t look like a Pixar movie, but his ears are perfect, his stick is a stick, and it didn’t cost me two hours of my life to appreciate it. Maybe the real art is knowing when to log off.