Truace
Subscribe

Artists Can Design Memory, and AI Can Predict It

Can artists deliberately design artworks to be remembered-or forgotten? Across three experiments, we tested whether visual-memory outcomes can be shaped through artistic design and how they are predicted by artists, laypeople, and artificial intelligence. We launched a nationwide art contest in which artists submitted works intended to be memorable or forgettable. Using recognition tasks with adults based in the United States (119 from Amazon Mechanical Turk, 892 from Prolific, and 62 visitors from a Chicago gal…

By the Truvace desk1 min read1 sourcePeer-reviewedTRV-2026-1786· cite & verify
Artists Can Design Memory, and AI Can Predict It
The long gallery, Louvre by Whistler, James McNeill, 1834-1903. Public domain

In brief

Researchers ran three experiments built around a nationwide art contest in which artists submitted works meant to be either memorable or forgettable. Adults in the United States then completed recognition tasks, and the artists' works were compared against predictions from artists, laypeople, and the ResMem neural network.

The result suggests that computational models can capture what makes images stick in memory better than human intuition can, even for trained artists. The source reports no deployment or downstream use, so the practical consequences for art, design, or media remain untested here.

Main points

  1. Artists achieved their intended memory outcomes, but memorability was more readily increased than reduced.
  2. ResMem consistently outperformed both human observers and artists in predicting which artworks would be remembered.
  3. Data-driven analyses found psychologically interpretable features that explained nearly half the variance in memory.
  4. Those features aligned with ResMem's predictions but not with human judgments.

The gain

In three recognition experiments, the ResMem neural network predicted which artworks would be remembered more accurately than human observers and the artists who made them.

The rundown

The study recruited adults based in the United States for recognition tasks: 119 from Amazon Mechanical Turk, 892 from Prolific, and 62 visitors from a Chicago gallery. This mix of online and in-person participants tested whether results held across settings.

The artists' success was asymmetric. Making an artwork more memorable worked more readily than making one more forgettable, which indicates that memory outcomes can be shaped by design but not equally in both directions.

The authors conclude that visual memorability can be engineered by humans, though asymmetrically, and is better captured by computational models than by human judgment.

Sources

  1. Peer-reviewedPsychological Science2026-10-13

The debate