If spending time on the couch, binging fast food, drinking too much alcohol or not paying into your company pension is ruining your carefully laid plans for life, it might be time to have a conversation with your future self.
With time machines not readily available, researchers at the Massachusetts Institute of Technology (MIT) have developed an AI-powered chatbot that simulates a user’s past self and offers observations and valuable wisdom in the hope of encouraging people to think more today about who they want to be tomorrow.
By digitally de-aging profile photos so that younger users appear as wrinkled, grey-haired seniors, the chatbot generates plausible artificial memories and weaves a story about a successful life based on the user’s current aspirations.
“The goal is to encourage long-term changes in thinking and behavior,” says Pat Pataranuthapong, who works on the Future You project at the MIT Media Lab, “which may motivate people to make smarter choices in the present that optimize their long-term well-being and life outcomes.”
In one conversation, an aspiring biology teacher asked a chatbot, a 60-year-old version of herself, about the most rewarding moment in her career so far. The chatbot, responding that she was a retired biology teacher in Boston, recalled a special moment when she turned a struggling student’s grades around. “It was so gratifying to see my student’s face light up with pride and accomplishment,” the chatbot said.
To interact with the chatbot, users are first asked to answer a series of questions about themselves, their friends and family, the past experiences that have shaped them, and the ideal life they envision for themselves in the future. They then upload a portrait image, which the program then digitally ages to create a portrait of them at 60 years old.
The program then feeds information from the user’s answers into a large language model to generate a rich synthetic memory for the simulated older version of itself, ensuring that the chatbot draws on a coherent background story when responding to questions.
The final part of the system is the chatbot itself, powered by OpenAI’s GPT3.5, which introduces itself as a potential older version of the user and can talk about their life experiences.
Pattaranuthapong has had several conversations with his “future self,” but the most memorable was when the chatbot reminded him that his parents won’t be together forever, so he should spend time with them while he still can. “The perspective I gained from that conversation is still influential to me today,” he said.
Users are told that their “future self” is not a prediction, but a potential future self based on the information they provide, and are encouraged to explore different futures by varying their survey answers.
be A preprint scientific paper on the projectA trial of 344 volunteers, which hasn’t been peer-reviewed, found that talking to a chatbot made people feel less anxious and more connected to their future selves. Pattaranthapong said this stronger connection should encourage better life choices, from focusing on specific goals and exercising regularly to eating healthier and saving for the future.
Ivo Vlaev, professor of behavioural science at the University of Warwick, said people often struggle to imagine themselves in the future, but doing so could lead to stronger adherence to education, healthier lifestyles and more careful financial planning.
He called the MIT project a “fascinating application” of behavioral science principles. “It embodies the idea of a nudge, a subtle intervention designed to steer behavior in a beneficial direction by making your future self more salient and relevant to the present,” he said. “Implemented effectively, this could have a profound impact on how people make decisions today with their future well-being in mind.”
“From a practical standpoint, its effectiveness will depend on how well it simulates meaningful, relevant conversations,” he added. “If users perceive the chatbot as authentic and insightful, it can have a significant impact on behavior. But if the interaction feels superficial or quirky, its impact may be limited.”
Source: www.theguardian.com
