Have you ever wanted to travel through time and see what you’ll look like in the future? Thanks to the power of generative AI, it’s possible.
Researchers at MIT and elsewhere have created a system that allows users to have text-based conversations online with AI-generated simulations of their potential future selves.
The system, named ‘Future You’, aims to help young people improve their sense of self-continuity in the future. This is a psychological concept that describes how connected a person is to their future self.
Research shows that a strong sense of future self-sustainability positively influences people’s long-term decision-making, from the likelihood of contributing to financial savings to the ability to focus on achieving academic success. Possibly.
Future You utilizes a large-scale language model that generates a relatable virtual version of a 60-year-old individual based on information provided by the user. This simulated future self can answer questions about what someone’s future life will be like. Provide advice and insight on the path they take.
In an initial user study, researchers found that after interacting with Future You for about 30 minutes, people reported feeling less anxious and felt a stronger connection to their future selves.
“Although we don’t yet have a real-time machine, AI could become a kind of virtual time machine. This simulation could help people think more about the consequences of the choices they make today. You can,” says Pat Pataranutaporn, a recent Media Lab doctoral graduate who is actively developing a program at MIT that advances human-AI interaction research. -Lead author of the paper on Future You.
Pataranutaporn is also co-lead author Kavin Winson, a researcher at KASIKORN Labs, on the paper. and Peggy Yin, an undergraduate student at Harvard University. So are Auttasak Lapapirojn and Pichayoot Ouppaphan of KASIKORN Labs. and senior author Monchai Lertsutthiwong (Head of AI Research at KASIKORN Business-Technology Group). Patti Mace, Germeshausen Professor of Media, Arts, and Sciences and director of MIT’s Fluid Interfaces Group, and Hal Hirsch, professor of marketing, behavioral decision making, and psychology at the University of California, Los Angeles. This is Mr. Field. The research will be presented at the IEEE Conference on Frontiers in Education.
realistic simulation
Research on conceptualizing one’s future self dates back to at least the 1960s. One of the earliest methods aimed at improving future self-continuity was having people write letters to their future selves. More recently, researchers have utilized virtual reality goggles to help people visualize themselves in the future.
However, none of these methods were interactive and had limited impact on users.
With the advent of generative AI and large-scale language models like ChatGPT, researchers have the opportunity to create simulated future selves that can discuss someone’s real-life goals and aspirations during normal conversation. I thought there was.
“This system makes the simulation very realistic. ‘Future You’ is much more detailed than what people can come up with just by imagining themselves in the future,” says Mace.
Users start by answering a series of questions about their current life, what’s important to them, and their future goals.
The AI system uses this information to create what researchers call a “memory of your future self.” This provides a backstory for the model to draw upon when interacting with the user.
For example, a chatbot can talk about the highlights of someone’s future career or answer questions about how a user overcame a particular challenge. This is possible because ChatGPT is trained on extensive data, including people talking about their lives, their careers, and their good and bad experiences.
Users utilize this tool in two ways. One is introspection, when you think about your life and goals as you build your future self, and the other is reminiscence, when you reflect on whether the simulation reflects who you want to be, Yin said. Masu.
“You can imagine Future You as a story search space, a chance to hear how some of the experiences that still touch us are metabolized over time,” she says.
To help people envision their future selves, the system generates older photos of users. The chatbot is designed to provide vivid answers using phrases like “When I was your age,” so the simulation feels like a real-life future version of the individual. .
The ability to accept advice from an older version of yourself, rather than a typical AI, could have a stronger positive impact on users considering an uncertain future, Hirshfield says.
“The platform’s interactive and vivid components provide users with an anchor point, making what could otherwise lead to anxious rumination become more tangible and productive,” he added.
But that realism can backfire if the simulation goes in a negative direction. To prevent this, Future You displays only one potential version of your future self and alerts you that you have the power to change your life. Providing alternative answers to a survey creates a completely different conversation.
“This is not a prophecy, but rather a possibility,” says Pataranutaporn.
help self-development
To evaluate Future You, we conducted a user survey with 344 people. Some users interacted with the system for 10 to 30 minutes, while others simply interacted with a generic chatbot or filled out a survey.
Participants who used Future You were able to develop a closer relationship with their ideal future self based on statistical analysis of their responses. These users also reported feeling less anxious about the future after the interaction. Additionally, Future You users stated that the conversations felt sincere and that their values and beliefs appeared consistent with the simulated future identity.
“This work blazes new trails by using established psychological techniques to visualize the coming era, your future self, with state-of-the-art AI. “This is exactly the type of research that academics should be focusing on, because model-building technology is fused with large-scale language models,” said Jeremy Bailenson, Thomas More Stork Professor at Stanford University, who was not involved in the study. says.
Building on the results of this initial user study, the researchers continue to fine-tune the way they establish context and key users to enable conversations that help create a stronger sense of self-continuity in the future.
“We want to guide users to talk about specific topics, rather than asking their future selves who the next president will be,” says Pataranutaporn.
They have also added safeguards to prevent people from abusing the system. For example, you could imagine a company creating a “future you” of a potential customer who achieved some great accomplishment in life by purchasing a particular product.
In the future, the researchers hope to explore specific applications of Future You, such as helping people explore different careers and visualize how their everyday choices affect climate change. I’m thinking.
They are also collecting data from the Future You pilot to better understand how people are using the system.
“We don’t want people to become dependent on this tool; rather, we want it to be a meaningful experience that will help them see themselves and the world differently and help them develop themselves. I hope so,” Mays said.
The researchers are grateful for the support of Thanawit Prasongponchai, a designer at KBTG and visiting researcher at the Media Lab.
