Artist uses AI to talk to your inner child

If we could talk to younger versions of ourselves, what would we say, what advice would we give, and how would we feel? Well, an artist had a similar idea after creating a chat with a artificial intelligence to talk to a younger version of herself — something like a “inner child“, training her to learn what she was like based on a diary written by herself as a young girl. She just takes a look.

AI that goes back in time?

Michelle Huang used his diaries, which were filled in over 10 years of his life, as a basis for feeding the artificial intelligence and combined it with the language model “OpenAI Generative Pre-trained Transformer 3 (GPT-3)" or "OpenAI GPT-3 Playground“, to achieve this feat. Among the things she wrote in her diary, much was said about dreams, fears and some secrets. She told users on Twitter that she created the AI ​​system so she "could have a real-time dialogue with her inner child."

O OpenAI GPT-3 Playground It has some rules to be used. There is an upper limit on how much content you can put in, so snippets of about 40 entries from ages 7 to 18 were selected, but Michelle ended up focusing on ages 10 to 14. After putting all the snippets she wanted into the tool, the artist claims he got responses that actually seemed to be something similar to how he thinks he would have responded during that time.

One of the questions was if she was happy with where the adult version was or if she was disappointed, so her younger self replied:

Honestly, I'm proud of you for everything you've accomplished. It hasn't been easy and I know you've made a lot of sacrifices to get to where you are. I think you are doing an amazing job and I hope you continue to chase your dreams and make a difference in the world.

artificial intelligence answer

Author's impressions

Responsible for this creation, Michelle Huang, still made several records about her relationship with the responses obtained by the tool:

When I told her that she was loved and cared for and safe: the words my past self always wanted to hear it felt like I was reaching into the past and giving her a giant hug, and I felt it come back to the present. When I asked her to write me a letter in the present day upon reading this, I felt the rumination spirals — the ones I sometimes fall into when I feel ashamed or disappointed — melt away a bit.

How Michelle felt in some moments of interaction

The artist also completed by implying that the tool improvement can bring many benefits, including therapeutic nuances:

These interactions really elucidated the healing potential of this medium: of being able to send love back to the past as well as receive love from a younger self. Overall this was a very mind-blowing but also a strangely affirming/healing experience that I didn't know I had access to using real data from my past self allowing me to connect with her in deeper and more tangible ways than usual. .

Michelle follows her reasoning about the advantages of this tool

How to make your inner child

Michele Huang created a series of tweets — thread — explaining how to make your own inner child using artificial intelligence. See below:

• Step 1: Have some material as a database. This can range from diaries, personal narratives, any type of written content that showcases your personality, voice, values, etc. If you don't have journals to reference, archived chats should work too;
• Step 2: after having some material collected access the GPT-3 playground🇧🇷 This site is basically where you can add your prompts, journal entries and define training parameters for your model. You will need to have an account opened on the site, as well as a payment method to run the model;
• Step 3: next you will start creating prompts in the playground. Michelle, for example, used a open artificial intelligence document🇧🇷 Something like “The following conversation pertains to YOUR NAME (CURRENT AGE) and YOUR YOUNGER NAME (YOUNGEST AGE). YOUR YOUNGEST NAME wrote these sentences [insert your records], while you currently ask [insert your question].

There are even more options like adding lightness or kindness to the responses so that you are not surprised by any possible bad manners on the part of the tool. The instructions show a real way to use artificial intelligence using programming codes.

And you, what did you think of the news? Tell us in the comments!

See also:

Google issues “red alert” after ChatGPT.

Pig iron: Business Insider e Daily Mail.