top of page

Subscribe to Our Newsletter

Thanks for submitting!

What Happens to Your Style When Every Reference Comes From the Same Feed?

12m
4 min read

The images we study shape the ideas we make. As recommendation feeds and AI assistants put more examples within reach, I wonder how often we're getting a wider view and how often we're looking at variations of what we already know.


I keep thinking about how much of creative work we learn just by looking. Before you ever open Photoshop or set up a light, you already have pictures in your head. You’ve seen how someone framed a portrait, used a particular shade of green or made a title sit across a photograph. Then you start putting your own version together.


Now think about where those pictures come from. A lot of the time, it’s the same few apps. You follow somebody whose work you like. The app shows you more people who make similar work. You save a few references, and pretty soon you have a whole folder of images that fit together.


That can be useful. I’ve also been wondering what we leave out of that folder.


The first example has a way of sticking around


Design researchers have been studying a problem called design fixation for decades. In a 1991 paper, David Jansson and Steven Smith showed that examples presented during a design exercise could influence the features people carried into their own solutions. More recent research has found a complication: designers can become attached to their own first ideas, even when nobody shows them an outside example.


That second finding matters. We can’t point at an algorithm and blame it for every familiar-looking poster or photograph. There are habits in the way we think, too. Once we find a composition that works, it’s tempting to adjust the colors and call the next version a new idea.


The research deals largely with design and engineering exercises. It doesn't measure the effect of a particular social media feed on a photographer's style. Still, it gives me a useful question to ask when I'm looking at a set of references: How many different possibilities did I actually consider?



Being shown more isn't always the same as seeing more


Generative AI adds another layer. A tool can give us several starting points in seconds. Sometimes that helps someone try an idea they might never have attempted on their own. There is evidence of that benefit, although the studies cover specific kinds of work.


In a 2024 Science Advances experiment, Anil Doshi and Oliver Hauser asked 293 people to write short stories. Participants who could request AI-generated ideas produced stories that outside readers rated more favorably on certain creativity measures, particularly among writers with lower starting scores. The AI-assisted stories were also more similar to one another.


That study was about eight-sentence fiction, with fixed prompts and limited interaction. It's an interesting result, but we shouldn't pretend it tells us exactly what happens to a portrait photographer, a costume designer or a student working in Illustrator.


What catches my attention is that two things can be true at once: more people may reach a better first result, and the collection of results may begin to look more alike. The question of who chooses the references becomes important.



Looking outside the frame


Imagine two portraits that are both well made. One uses warm light to draw attention to a face. Another is in black and white, with architectural shapes interrupting the view. You could describe them using the same camera vocabulary, but they ask you to notice different things. A useful reference library should make room for those differences.


A seated woman holds white flowers in warm light against a dark brown background.
Portfolio photograph by Theoplis Stewart II. A seated woman holds white flowers in warm light against a dark background. Here, the face and flowers draw attention through contrast and color.

Black-and-white photograph of two people seen behind a repeating white balustrade.
Portfolio photograph by Theoplis Stewart II. Two people appear behind a repeating balustrade in black and white. The foreground changes how the viewer finds the subjects.

I want to keep looking beyond whatever an app thinks I'll click next. A photograph in a local archive, a hand-painted sign, a film I wouldn't ordinarily watch, the way light passes through a window or an old piece of packaging can all teach something different. None of those sources is automatically better. What matters is being able to explain what you noticed and why it belongs in your work.


That also means learning about the people and places behind a reference. Pulling an image from a museum website doesn't tell you everything about its history or the terms under which it can be reproduced. The Library of Congress distinguishes material with no known publication restrictions from material whose rights have not been evaluated. I think that's a useful reminder to slow down long enough to understand a source.



A small experiment worth trying


Here's an exercise I would like to try. It hasn't been conducted for this article. Choose one assignment, perhaps a poster or portrait mood board. Gather five references from a familiar feed, then five from places that require a different kind of search: a book, a physical location, an archive, a painting or a conversation with another creative. Note what each image actually contributes.


Make two early concepts from the same brief. Put them beside each other without labeling the reference group. Ask someone which differences they notice. Then go back and ask yourself a harder question: which choices can you defend without pointing to the reference and saying, "I liked the way it looked"?


It wouldn't prove that feeds are hurting creativity. One small experiment can't settle that. It could show me where I'm repeating habits and where I'm beginning to see something new.


We're going to keep learning new software. I'm interested in what happens when we also put some energy into learning where our ideas come from. If we get better at asking for an image, I'd like us to get just as good at knowing why we're asking for that particular one.


Cover photograph: Theoplis Stewart II, from the author's portfolio.

Recent Posts

See All

Comments


Join our mailing list

©2024-2026 by Theoplis Stewart II.

bottom of page