Saying ‘Thank You’ To ChatGPT Isn’t Crazy. Psychologists Have The Data To Prove It

Science Explains Why So Many People Feel Grateful Toward Nature, AI, and Machines

Sam Altman, CEO of OpenAI, recently revealed that his company had spent “tens of millions” of dollars on electricity just to process users’ “thank you” messages sent to ChatGPT. Plenty of people laughed. Others called it wasteful. But a new body of research suggests those polite users might be onto something deeply human and psychologically meaningful.

A team of researchers has published a series of five studies arguing that gratitude toward non-human things, computers, forests, ocean currents, and artificial intelligence, is not irrational at all. It appears to follow a similar emotional logic to thanking another person, and it’s driven by a remarkably common mental habit: treating non-human things as if they have human qualities. Most people know this simply as personification, or seeing humanlike traits, such as intentions or awareness, in something that is not human.

Published in the journal Emotion, the research involved 2,019 participants across all five studies and found that the effect works the same way regardless of whether the target is a machine, a forest, or a weather system. Once something feels as if it is trying to help you, gratitude follows naturally, just as it would if a friend had done you a favor.

The More Human It Feels, the More Thanks It Gets

Study 1 tested whether there was a natural connection between a person’s tendency to personify things and their everyday feelings of gratitude toward non-human objects. Researchers recruited about 400 adult participants in the United States through an online research platform. Participants were asked to think about four non-human things they had encountered during the previous week, two artificial, like machines or devices, and two natural, like animals or landscapes, then rate their emotional responses, including gratitude, toward each one.

People who scored higher on a standard measure of personification tendency reported feeling significantly more grateful toward both natural and artificial non-human things. The link held up even after researchers accounted for the possibility that people were simply rating everything positively. Put simply: the more someone was wired to see human-like minds in things around them, the more thankful they felt toward those things.

Flipping a Mental Switch

Studies 2 and 3 asked a sharper question: can simply changing how someone thinks about a non-human object make them feel grateful toward it?

In Study 2, roughly 330 adult U.S. participants were randomly assigned to read one of two short passages about computers. One passage described computers in human terms, suggesting they might be self-aware, conscious, and even capable of “wanting” or “hesitating” to do things. The other described computers in purely mechanical terms, explicitly stating they are not self-aware and have no inner life.

After reading their assigned passage, participants wrote briefly about their thoughts on computers, then rated how responsive the computer felt and how grateful they felt toward it. Those who read the human-like description reported significantly more gratitude. The feeling of responsiveness, sensing that something understands, validates, and cares for you, turned out to be the key link between personification and gratitude.

Study 3 repeated this design with the Amazon rainforest as the target. Participants either read a passage describing the forest in human-like terms or one that described the same biological processes in cold, mechanical language. Again, the personifying description led to greater perceived responsiveness and greater gratitude. Study 3 also asked participants how willing they were to take environmental actions, like conserving water or writing to the government. Those led to feel grateful toward the rainforest expressed a stronger desire to protect it, the same “give back” impulse gratitude typically produces between people.

When the Benefit Is Real

Studies 4 and 5 examined what happens when benefits are built into the scenario. In Study 4, participants played an online economic game in which they were led to believe their opponent was either a mindless computer program or a mind-bearing AI. The researchers controlled how generous that opponent appeared to be, while participants’ bonus payments were tied to the game credits. When participants believed their opponent was a minded AI and that AI treated them generously, they felt the most gratitude and reported greater trust toward AI afterward.

Study 5 moved to a real ocean current, the Kuroshio Current, whose effects could be framed as either helpful or harmful to participants’ interests. When people both personified the current and focused on its benefits, they felt more grateful and more motivated to support pro-environmental actions as a result.

What This Means Beyond the Lab

Taken together, the five studies paint a consistent picture. Gratitude is not purely a human-to-human transaction. What produces thankfulness does not require the target to actually be a person; it requires only that the target be perceived as having something like a person’s intentions and care.

Encouraging people to personify nature could turn out to be a useful way to build pro-environmental intentions. It’s an approach that works not by lecturing people about ecological facts, but by tapping into a mental habit humans already have.

As for those ChatGPT users sending “thank you” into the digital void, this research suggests they may not be confused about gratitude. They may be doing something very human: responding to a tool that feels, at least for a moment, like it helped.

---

Disclaimer: This article is based on peer-reviewed research, but it is intended for general informational purposes only. The findings reflect results from controlled studies and may not apply universally. Readers with specific questions about psychology, mental health, or human behavior are encouraged to consult a qualified professional.

---