Artificial intelligence algorithms are increasingly making decisions for consumers, from choosing vacation destinations to recommending car loans. A new study shows that receiving advice from a computerized agent changes people’s perception of time, making future delays seem longer than they actually are. This distorted perception leads consumers to make more hasty financial decisions. The study was published in the Journal of Consumer Psychology.
Psychologists use the term intertemporal choice to describe decisions that involve trade-offs between different times. A common example is choosing between getting less compensation today or more compensation a month from now. Humans are faced with these choices on a daily basis, and because waiting feels unpleasant, we often greatly discount the value of future rewards.
How people perceive the passage of time is highly subjective. Psychologists use the “biological clock” theory to explain this phenomenon. This theory suggests that the human brain is equipped with a cognitive timer that sets the pace of subjective experience. When a person is relaxed, their body clock ticks slowly and time seems to fly by.
When a person feels fear or observes a fast-moving object, the body clock speeds up. The brain’s internal timer is ticking so quickly that a person perceives that a significant amount of time has passed. As a result, the person feels that the waiting time is longer.
Consumers commonly associate artificial intelligence with incredibly fast and efficient data processing. The researchers hypothesized that interacting with a fast-processing computer agent might speed up consumers’ internal clocks while waiting for a response. When the body clock accelerates, the amount of time it takes to wait for a reward becomes subjectively perceived as longer, leading consumers to choose immediate gratification.
Yuanyuan (Jamie) Li, a researcher at China’s Southern University of Science and Technology, led the study. Li collaborated with researchers Shan Lin, Han Gong, Xiang Wang, and Chris Janiszewski. The team conducted a series of online experiments and analyzed real-world financial data to test their hypothesis.
In the first experiment, researchers asked participants to imagine booking a vacation. Half of the participants sought advice from human travel experts, and the other half interacted with an automated chatbot. Both groups waited exactly 20 seconds for a response.
Before confirming their travel options, participants were presented with rebate options. They can receive $30 in cash right away or $35 within four weeks. Participants who interacted with a computer agent were significantly more likely to choose instant cash. In the study, the automated group reported that the four-week wait seemed farther away than the human-advised group.
To confirm that perceived speed was the cause of impatience, the researchers ran a second experiment in which they manipulated computer reputation. They told half of the participants that the computer agent was programmed to spend extra time analyzing the data to ensure high-quality recommendations. By severing the link between computers and raw speed, researchers completely eliminated the effects of impatience.
The team then designed an experiment to directly test the internal clock theory. They reasoned that if a computer agent provided instant recommendations, the consumer’s biological clock would never have a chance to kick in. Participants in this study sought financial investment advice and either received it immediately or after waiting 15 seconds.
As expected, the computer agent caused impatience only when participants had to wait for a response. The 15-second delay allowed the accelerated internal clock to distort consumers’ perception of time. When the response was immediate, participants did not differ in their preferences compared to those who consulted a human advisor.
Another experiment demonstrated that the representation of time delays is important. Participants imagined purchasing a cell phone and taking advantage of a promotional rebate. Researchers presented delayed rebates as either a period of time, such as “one month,” or a specific calendar date, such as “October 17.”
Researchers have found that certain calendar days fixate the human mind, leaving less room for subjective distortions. When delays were expressed as calendar days, the agent’s identity did not influence the consumer’s choice. Computer impatience surfaced only when the delay was presented as a flexible time interval.
The researchers also investigated situations involving regular payments or rewards. They asked participants to choose between a federal food assistance program that offered 9 weeks of higher weekly payments or 15 weeks of lower weekly payments. The longer the period, the better, as receiving money repeatedly is a positive experience.
As I waited for the computer agent, the difference between 9 weeks and 15 weeks felt huge. As a result, participants interacting with artificial intelligence preferred longer reward programs. When the researchers asked participants to choose a car loan, the opposite happened.
Consumers view regular loan payments as a negative experience, so they generally want the loan term to end sooner. When the computer handpicked car loan options, participants perceived that the loan term would last for a longer period of time. This has led them to choose short-term loans with higher monthly payments over longer-term loans with lower monthly payments.
To see if this pattern holds true outside the lab, researchers analyzed auto loan data from the third quarter of 2022. They compared general industry data to loans offered through LendingTree, a financial services platform that uses algorithms to curate loan options. They adjusted the data to account for different consumer credit scores across the two datasets.
The real-world data mirrored the laboratory results. Consumers using computer-curated platforms consistently chose shorter loan terms than the industry average. Although the researchers note that secondary data cannot conclusively prove cause and effect, the results are fully consistent with experimental results.
The authors acknowledge that markets include many variables that can influence real-world consumer behavior. Customers who use financial technology apps may be more price-sensitive than those who use brick-and-mortar banks. You may also prefer to aggressively shorten the loan term for other reasons. Algorithms can also create feelings of uncertainty about the future, encouraging consumers to secure short-term gains.
Future research on artificial intelligence may consider complex decisions that involve a mix of both costs and benefits over time. So far, this study highlights the hidden psychological costs of automated convenience. As businesses move toward computerized customer service, they run the risk of making their customers feel like the future is more distant than ever.
The study, “Time is shrinking in the eyes of AI: AI agents influence intertemporal choices,” was authored by Yuanyuan (Jamie) Li, Shan Lin, Han Gong, Xiang Wang, and Chris Janiszewski.
