
A persistent challenge in personal finance is getting individuals to save adequately for retirement. A key reason for this difficulty is that people prioritize immediate rewards over long-term benefits. When retirement is decades away, the consequences of saving (or failing to save) feel abstract and distant, making it hard for individuals to take action.
A behavior AI tool, The Retirement Visualizer™, was co-invented with Joseph Reiff, assistant professor of marketing, Dan Goldstein, senior principal research manager at Microsoft Research NYC, Hal Hershfield, professor of marketing and behavioral decision-making and Marketing Area Chair at UCLA, and Shlomo Benartzi, professor emeritus at UCLA, helping individuals overcome present bias and increase their motivation to save for retirement.
“With the rising tide of AI, we've become interested in how we can leverage these tools to heighten people's motivation to save,” says Reiff.
One of the main challenges in financial planning is that the benefits of saving are distant and abstract. To address this, the team developed the Retirement Visualizer™ using OpenAI’s GPT and DALL·E, designed to make the future feel more tangible.
"When someone inputs their desired retirement, say, ‘I want to be on a cruise with my friends in the Caribbean,’ the GPT will create a first-person visual of that desired retirement using their DALL·E function," Reiff explains. "It creates a photorealistic image of what the person wrote for their goal."
To test the tool, the team is currently partnering with financial service providers to design large-scale randomized controlled trials. Participants will be divided into two groups: the control group writes about their retirement goals and then decides on their savings rate, and the treatment group follows the same process but receives AI-generated images of their envisioned retirement before making their savings decision. The prediction is that in the treatment condition, the AI tool can help individuals vividly imagine what they can get out of saving for retirement.
As a result, they will save more than people in the control condition who don't see those visuals.
By making retirement more vivid and personally meaningful, the approach may encourage higher savings rates and serve as a scalable intervention that financial institutions can integrate into retirement planning tools.
As Reiff puts it, "We're really excited about this. I think it would contribute to a growing body of work showing how these tools around generative AI can be used to influence people's behavior in beneficial ways.”
The Retirement Visualizer™ was covered by PLANADVISER’s national conference last year.
Media Contact
Greg Muraski
Media Relations Manager
301-405-5283
301-892-0973 Mobile
gmuraski@umd.edu
Get Smith Brain Trust Delivered To Your Inbox Every Week
Business moves fast in the 21st century. Stay one step ahead with bite-sized business insights from the Smith School's world-class faculty.