Why your future self feels like a stranger
Psychologist Hal Hershfield, now at UCLA Anderson School of Management, has spent years studying what he calls future self-continuity: how connected you feel to the person you will be. In a 2011 review he argued that many people think of their future self almost as a different person, and that those who feel more connected tend to make choices that favour the long term, such as saving.
In one set of experiments, published in 2011 with colleagues at Stanford, people who saw age-progressed images of their own faces in virtual reality were more willing to accept later rewards over immediate ones. His 2023 book, Your Future Self, turns this work into practical exercises, including writing letters.
Why a letter from the future works
Psychologists Yuta Chishima and Anne Wilson tested a two-way version of the exercise with Japanese students, published in 2020. Students who wrote to their future self and then wrote a reply as that future self felt more connected to their future, and a month later reported more career planning. Students who only wrote to their future self did not show the same benefits.
In a 2021 online experiment with 738 adults in Japan during the early COVID-19 pandemic, Chishima, Liu and Wilson found that both writing to the future self and writing from it reduced negative feelings straight away, compared with a control group.
These studies used letters people wrote themselves. They show that imagining a future perspective can help; they do not show that an AI-written letter has the same effect.
What AI adds, according to early studies
Two recent studies tested AI versions of the future self.
- Future You (MIT Media Lab, 2024). Pat Pataranutaporn and colleagues, including Pattie Maes and Hal Hershfield, built a chatbot that plays your future self at age 60, using an age-progressed photo and a backstory generated from your survey answers. In the latest version of the preprint, a randomised study with 344 participants aged 18 to 30 found that one short chat reduced anxiety and increased future self-continuity compared with controls. Measures such as optimism and sense of agency did not change significantly, and the authors warn about over-reliance and misleading pictures of the future.
- Letters from Future Self (CHI 2025). A one-week study with 36 young adults compared writing your own reply with receiving letters or chat from an AI future self. The AI raised engagement, but the overall benefits for future orientation and career thinking were comparable to writing the reply yourself.
The honest summary: AI can make the exercise easier to start and more engaging. The value seems to come from imagining your future perspective, which you can do with or without AI.
The exercise, step by step
- Write to your future self firstSpend ten minutes writing a short letter to yourself in ten or twenty years: what you are working on, what worries you, what you hope for.
- Describe your present honestlyGive the AI a few sentences about your life now and the letter you wrote. Leave out names, addresses and anything you would not want stored by the provider.
- Ask for a draftUse the prompt below, or your own version. Ask for a letter, not advice.
- Make it yoursEdit the draft until it sounds like you and feels possible. Delete anything that is flattering but empty.
- Write the reply yourselfWithout AI, answer the letter from the point of view of your future self. This is the step the research supports most directly.
- Keep it and revisitSave both letters with the date. Reread them in six months and notice what has changed.
A prompt you can adapt
Copy this into any chatbot, replacing the parts in brackets:
“I am [age] and I [two or three sentences about your life now]. In ten years I hope to [your hopes]. My biggest worry is [worry]. Write a short, warm letter to me from myself in ten years. Be specific and realistic, not perfect: mention one thing that turned out harder than expected and one small habit that made a difference. Do not predict exact events, and do not just tell me what I want to hear.”
Useful follow-ups: “Rewrite it in plainer words, the way I talk.” “What would my future self say about the decision I am facing now?” “What would they thank me for doing this year?”
Limits to keep in mind
- It is imagination, not prediction. No AI knows your future. Treat the letter as one possible story, not a forecast.
- It may flatter you. Chatbots tend to agree with the user. Ask for honesty and cut sentences that feel like a greeting card.
- It reflects what you give it. A letter built from a gloomy description will lean gloomy; one built from wishful thinking will lean unrealistic.
- It is not therapy. If thinking about the future brings up anxiety or low mood, talk to someone you trust or a professional. Our wellbeing guidance lists free helplines.
Using a clone of yourself
A general chatbot writes in a generic voice. If you want the letter to sound like you, a clone built from your own messages can draft it in your usual style. Keep in mind that such a clone knows your past, not your future, so you still need to tell it your hopes.
Memory Clone is a Windows desktop app, currently in alpha, that builds an AI simulation of you from chats you export yourself and your notes. Imported originals stay on your PC, cloud processing is off until you switch it on for a clone, and your content is not used to train models. Read how to clone yourself with AI for the options, text your past self for the backward-looking version of this exercise, and AI journaling with your own messages for a regular routine.
Frequently asked questions
Is it better to write a letter to or from your future self?
In Chishima and Wilson’s student study, doing both, writing to your future self and then replying as them, produced benefits that writing one letter alone did not.
How far into the future should the letter come from?
Studies have used horizons from a few years to several decades; Future You used age 60. Pick a distance that feels meaningful but still imaginable for you.
Does talking to an AI future self reduce anxiety?
One randomised study, Future You, reported lower anxiety after a single short chat. That is promising but early evidence, measured soon after the session.
Should I let the AI write the whole letter?
Use it for a first draft at most. Early research suggests the benefit comes from imagining your future perspective, so editing the letter and writing your own reply matter more.
Is it safe to share my goals with a chatbot?
Anything you type is sent to the provider. Leave out names, health details and other sensitive information, and check the service’s privacy settings.
Sources
Checked on . Third-party products and help pages change often; follow the links for current details.
- Hershfield, “Future self-continuity: how conceptions of the future self transform intertemporal choice” (Annals of the New York Academy of Sciences, 2011) (PubMed Central)
- Hershfield et al., “Increasing saving behavior through age-progressed renderings of the future self” (Journal of Marketing Research, 2011) (Stanford Virtual Human Interaction Lab)
- Hershfield, Your Future Self: How to Make Tomorrow Better Today (2023) (Little, Brown Spark)
- Chishima & Wilson, “Conversation with a future self: A letter-exchange exercise enhances student self-continuity, career planning, and academic thinking” (Self and Identity, 2020) (Taylor & Francis)
- Chishima & Wilson, What would your future self tell you? (2020) (Society for Personality and Social Psychology)
- Chishima, Liu & Wilson, “Temporal distancing during the COVID-19 pandemic: Letter writing with future self can mitigate negative affect” (Applied Psychology: Health and Well-Being, 2021) (Wiley)
- Pataranutaporn et al., “Future You: A Conversation with an AI-Generated Future Self Reduces Anxiety, Negative Emotions, and Increases Future Self-Continuity” (2024) (arXiv)
- Future You project overview (MIT Media Lab)
- Jeon et al., “Letters from Future Self: Augmenting the Letter-Exchange Exercise with LLM-based Agents to Enhance Young Adults’ Career Exploration” (CHI 2025) (arXiv)
This article is general information, not medical, legal or financial advice. Memory Clone creates an AI simulation from the material you provide; it is not the person.