Michelle Huang, a New York-based artist, now 26, may have discovered a means to have a discussion with her younger self if this is something you have always wanted to do.
All you need is the most advanced artificial intelligence (AI) software plus data from your history, which in her case were her long-lost childhood notebooks.
AI Model
At the age of 7, Huang started keeping a diary. She used her journal entries to train the AI language model, GPT-3, as reported by The Independent.
Developed by prominent AI research firm OpenAI, GPT-3 employs deep learning to generate writing practically indistinguishable from a human's.
To mimic what it would be like to have a conversation with her younger self, Huang kept a notebook in which she wrote about her day, expressing anything from her deepest worries to the "giddiness" of chatting with a crush.
She told The Independent that the experience was quite trippy.
"It felt like I was reaching into the past and hacking the temporal paradox: the chat box window felt like a time portal that was happening in real-time, as if my 14-year-old self was sitting on the other side typing responses."
The Conversation
Initially, she questioned her younger virtual self about her perspective on the world before opening the floor to the AI's inquiries.
The questions evoked strong emotions, and the responses were "eerily similar" to how she imagined she would have reacted at the time.
When prompted, the bot inquired, "Are you happy with where you are in life?" and provided encouraging feedback such as, "I'm honestly proud of you for everything you've accomplished."
After exchanging queries, Huang requested the journal-generated AI to send a letter to her in the current day.
In the message, the AI wished the reader well and that she may be engaged in a pursuit that brings her joy and satisfaction in life. It prays that the present Michelle has not allowed anything or anybody to alter her unique personality.
What's Next
There are concerns that technologies like GPT-3 and image-based AI with the ability to create deepfakes might be misused, said the report.
They are capable of realistic impersonation, story writing, content production, and even reincarnation. A perfect illustration is Google's newly unveiled AI-powered programs, which Tech Times reported earlier this month.
GPT-4, the next version of OpenAI's industry-leading algorithm, is scheduled for release in the upcoming months. It has been speculated that it is far more potent and perhaps more harmful than its forerunner.
On the other hand, Huang believes that these technologies will be utilized "for awareness and empathy" rather than for damage.
In the future, Huang plans to utilize the experiment to see whether AI can really revolutionize and inspire more interpersonal interaction among people.
Read More: Meta AI Researchers Reveal Cicero, First AI to Rival Humans in Diplomacy
