The Audit: Are You Shooting Like AI Before AI Arrived?

Paul Delaroche was a successful painter when he saw the daguerreotype demonstrated in 1839. He walked out convinced that his profession was over. He was probably wrong—painting survived by discovering what it was actually for. Now photographers are having the same panic. And the uncomfortable question underneath: were we already shooting like AI before AI showed up?

The story goes that Delaroche walked out and said "From today, painting is dead." But historians can't find a reliable source. The quote was probably invented. It appears to have been embellished decades later. What we know is that Delaroche did see early daguerreotypes, did understand the implications, and did eventually embrace photography rather than reject it. But the quote survived anyway—because it was true enough. Because it pressed on a pain point. Because it named something real: the terror of watching a machine do what you spent your life learning to do. In 2026, our version is: "From today, photography is dead." And like a lot of us, you've been thinking about that a lot.

What Actually Died When Photography Arrived

Painting didn't die. Painting's obligation to document died. The thing painting was paid for—to record the fact that something existed—that job went to the camera. And painting was forced, furious and disoriented and without a business model, to figure out what it was actually for. Monet didn't paint water lilies because he couldn't photograph them. He painted them because photography couldn't do what he was doing. He was painting perception itself. The way light doesn't illuminate things so much as create them. The way a surface shifts in the seconds you're looking at it. Impressionism, Surrealism, Cubism—the work we study now—didn't come from painters getting faster or cheaper. It came from painters with no choice but to ask a harder question: what does a brush do that light cannot?

The Uncomfortable Audit: Were You Already Shooting Like AI?

A sea of images from photographers—amateurs and pros alike, seen daily, sometimes hourly—rewired how we saw. Not just in comparison and self-doubt, though that was real. In a more systemic way. We stopped shooting for truth and started shooting for algorithmic legibility. For trends and hashtags. For the scroll. Overcast landscapes. Latte art on distressed wooden tables. Laundromats at night with neon signs. Moody portraits with shallow depth of field on flat gray skies. Safe compositions. Exaggerated negative space. Desaturated warm tones. Work that looked like it came from a discount art store. Professional-ish. Consumable. Forgettable. And the market rewarded us for it. We rewarded ourselves with likes. With follower counts. With convincing ourselves we were innovative. What we were actually doing was handing Instagram—and every other platform—billions of images to learn from. We built the training data. We formed the pattern. We fabricated the cage and then acted surprised when something climbed inside and started replicating the lock. AI didn't invent average. It just harvested the billions of images we posted and figured out what "quality" looks like. Now it outputs a varnished, blended version of all that generic professional-ish work back at us and we're supposed to compete with it.

The Commodity Work vs. The Work That Requires You

The photographers who built portfolios on that professional-ish work will be the first to feel it. Because that's exactly what AI has been training on. The fear you hear at conferences? That's recognition. It's the photographer who spent fifteen years building a "consistent" aesthetic realizing that consistency is now a twenty-dollar subscription. Here's the controversial take: was your work already that replaceable? Or did you make it that way? When you audit your own portfolio, the images that fail aren't technical failures. They're the ones where you showed up as a professional and left yourself outside. But when you look at work that has a pulse—at work that required you specifically to exist—that's the frame AI can't replicate. It's the photographer who decided to leave a light switch in the frame because a light switch next to power is a comment. It's the frame you almost didn't take. It's the version of you that was curious about the person in front of your camera—genuinely curious, not performing curiosity. AI cannot care about your subject's dinner plans. It cannot be drawn to the specific quality of light at 4 PM on a Tuesday in November. It cannot make the decision to sacrifice the brief because what's actually happening in the room matters more than what you were paid to capture.

Key Takeaways

• Painting survived the daguerreotype not by fighting it, but by discovering what a brush does that light cannot—its own irreplaceable function.
• The commodity work—work with no point of view, no specific human decision-making, pure technical execution—will be the first to be absorbed by AI, and that's not tragic, that's clarifying.
• We built the training data ourselves by chasing algorithmic legibility instead of truth for a decade; we fabricated the cage.
• The work that requires you—the work where your specific curiosity, your specific choices, your specific humanity is load-bearing—AI cannot replicate because AI doesn't care about anything.
• The audit isn't whether AI can replace us; it's whether there's enough of you in your portfolio that there's something to replicate.

The Terrible Take

The real work isn't fighting the technology. It's going back to the shoot where something went sideways and you stayed anyway. Go back to the frame you almost didn't take. Go back to the version of you that hadn't yet learned to pre-visualize the grid. That's where the work is. Not ahead of you. Behind you, waiting to be understood.