Dan Ariely on AI and Disposable Work: What Happens When Drafts Cost Nothing?

For most of modern working life, producing a first draft took effort, and that effort made people careful about what happened to it. Artificial intelligence has changed the arithmetic. In an October essay on his own website, behavioral economist Dan Ariely raises a question that few productivity discussions touch: what happens to meaning when work becomes fast, plentiful and easy to throw away?

Drafts That Cost Nothing

Ariely’s essay, “The Shredder Problem,” is mostly about the human cost of ignored work. He describes a university report that his research team wrote carefully and that drew only a note about mismatched identification numbers, after which the team became less willing to invest similar effort. His argument is that people sustain effort when they believe someone will engage with what they produce.

AI tools complicate that picture. When a draft takes seconds to generate and is discarded just as quickly, there is little attention left for any one output. Ariely suggests that such work may feel temporary to the people involved, a possibility that Programming Insider picked up in its coverage of the essay, under the headline about the hidden cost of making employees feel replaceable.

What the Research Does and Does Not Say

It is important to be precise about the evidence. Ariely’s experiments, conducted with Emir Kamenica and Drazen Prelec, did not involve AI. They showed that people built fewer Lego figures when they watched their work be dismantled, and quit a letter-finding task at higher pay per sheet when their sheets were ignored or shredded. A 2026 study in Psychological Science by Yilin Liu and Fan Yang found that most children as young as four preferred keeping their creations intact.

Applying that research to AI is an extension, not an established result. Ariely frames it as a concern worth watching, and coverage of his essay has been careful to say the same. The underlying logic is simple: if meaning depends on work being seen and used, then any workflow that generates outputs faster than anyone can attend to them may drain it.

Readers who want to follow how the argument develops can also find Ariely on Medium, though the essay itself remains the best place to start.

Dan Ariely’s Other Use for AI

Ariely has also floated a more constructive role for the technology. In a September interview with the Israeli outlet Mako, he suggested that AI chatbots could serve as a kind of standing devil’s advocate, a tool people turn to when they are in a disagreement, not to hand them an answer but to stress-test their own reasoning. It is a narrower use than most people associate with AI, and it points in the opposite direction from disposable output: it asks the person to defend or revise what they believe.

That idea connects to Dan Ariely’s earlier writing on democracy, which explored ways voting could give citizens more say. The connection is thematic, not causal: in both cases, the underlying concern is attention and consequence, whether a person’s contribution is taken seriously.

Ariely also has a presence on Instagram, but the longer essays are where the argument is made in full. The practical advice, in the meantime, is the same as ever: keep what people make where you can, and let them see that it was read.