‘Future You’ is an innovative AI tool that allows individuals to engage in a simulated dialogue with a version of their future selves. This chatbot aims to alleviate anxiety, enhance positive feelings, and assist users in making more beneficial daily decisions.
Have you ever wished you could time travel to glimpse your future self? Now, with the advancements in generative AI, you can do just that.
Researchers from MIT and other institutions have developed a platform where users can participate in a text conversation with a future version of themselves created by AI.
Known as Future You, this system focuses on helping young individuals strengthen their connection to their future selves, a psychological concept that explains the degree of relationship one feels with their future persona.
Studies indicate that a heightened sense of future self-continuity can positively affect long-term decision-making, influencing areas such as financial savings and academic achievement.
Future You employs a sophisticated language model that utilizes user-provided information to construct a relatable virtual version of the individual at age 60. This simulated future self can respond to inquiries about what life may hold in the future and provide guidance based on possible paths.
An initial study revealed that participants who engaged with Future You for about 30 minutes reported reduced anxiety and felt a stronger connection to their future selves.
“While we still lack a real time machine, AI can serve as a form of virtual time travel. This simulation encourages people to reflect on the ramifications of today’s choices,” explains Pat Pataranutaporn, a recent Media Lab doctoral graduate currently advancing human-AI interaction research at MIT and co-author of a paper on Future You.
Pataranutaporn collaborated on this paper with Kavin Winson, a researcher at KASIKORN Labs; Peggy Yin, an undergraduate at Harvard University; Auttasak Lapapirojn and Pichayoot Ouppaphan from KASIKORN Labs; as well as senior authors Monchai Lertsutthiwong, head of AI research at KASIKORN Business-Technology Group; Pattie Maes, Germeshausen Professor of Media, Arts, and Sciences leading the Fluid Interfaces group at MIT; and Hal Hershfield, a professor of marketing and decision-making at UCLA. Their research will be presented at the IEEE Conference on Frontiers in Education.
A realistic simulation
Research into how people visualize their future selves dates back to at least the 1960s. One early technique encouraged individuals to write letters to their future selves, while more recent methods have involved virtual reality technology to help users envision their future versions.
However, these earlier approaches lacked interactivity, limiting their effectiveness.
With the rise of generative AI and powerful models like ChatGPT, researchers recognized the opportunity to create a more engaging simulated future self that could actively discuss real personal goals and dreams.
“The system offers a highly realistic simulation. Future You provides much more detailed interactions than what someone could generate by merely imagining their future,” says Maes.
Users start by answering a series of questions about their current lives, values, and aspirations for the future.
The AI system then constructs what the researchers term “future self memories,” which serve as a backdrop for interactions with the user.
For example, the chatbot might recount significant moments in a person’s future career or address how the user managed to overcome specific challenges. This is possible due to ChatGPT’s extensive training on conversations about people’s lives, achievements, and experiences.
Users interact with the tool in two primary ways: through introspection, as they reflect on their aspirations while formulating their future selves, and through retrospection, assessing whether the simulation aligns with their ideal future identity, according to Yin.
“You can think of Future You as a narrative exploration space. It allows you to reframe emotionally significant experiences over time,” she states.
To aid users in visualizing their future selves, the system generates an age-advanced image of the user. The chatbot communicates in a relatable manner, using phrases like “when I was your age,” enhancing the realism of the future persona.
Receiving guidance from an older version of oneself, rather than a generic AI, can significantly impact individuals contemplating an uncertain future, comments Hershfield.
“The interactive and descriptive nature of the platform provides users with a grounding point, transforming anxious thoughts into actionable insights,” he adds.
Yet, this realism might lead to negative outcomes if the conversation veers toward unfavorable scenarios. To mitigate this risk, Future You is designed to remind users that it presents only one potential future and that they can influence their life path. Answering the initial questionnaire differently results in an entirely new conversation.
“This is not a prediction, but rather a possibility,” says Pataranutaporn.
Aiding self-development
To assess the efficacy of Future You, a study involved 344 participants. Some users engaged with the system for 10–30 minutes, while others interacted with a basic chatbot or solely completed surveys.
Those who utilized Future You reported a closer connection to their envisioned future selves, as evidenced by a statistical analysis of their responses. Moreover, these users experienced less anxiety regarding the future following their interactions, and they found the conversation sincere, with their future selves exhibiting consistent values and beliefs.
Building on the findings from this initial user study, researchers are refining their approach to establishing context and priming users for discussions that enhance future self-continuity.
“We aim to guide users in discussing pertinent topics rather than asking their future selves trivial questions,” Pataranutaporn notes.
Furthermore, safeguards are being developed to prevent misuse of the system. For example, a concern could arise if a company creates a “future you” that showcases a customer’s ideal outcome tied to a specific product.
In the future, the researchers hope to investigate specific applications for Future You, like assisting individuals in exploring various careers or understanding how their daily choices might affect climate change.
They are also gathering insights from the Future You pilot to gain a better understanding of user interactions with the platform.
“Our goal is not to make users reliant on this tool, but rather to offer a valuable experience that reshapes their perception of themselves and their surroundings, contributing to their personal growth,” Maes concludes.
The researchers would like to express their gratitude to Thanawit Prasongpongchai, a designer at KBTG and visiting scientist at the Media Lab, for their support.
