365 Days of AI: Experiment and Guidelines

This is a post that outlines the beginning of a new experiment in which I’ll make something with AI for 365 days.

Like many others, the recent advancements in artificial intelligence, more specifically, GPT technology, has captured my curiosity. I use OpenAI’s ChatGPT every day and I use it for many things. I am constantly asking it questions across every dimension of life. I ask it for feedback. I ask it to collaborate. I ask it to explain things to me. I ask it why I feel a certain way. I ask it to do things for me. Some of the things that I send to ChatGPT are messages and questions that I used to send to friends.

I have found my early GPT experiences to be kind of weird because they make computers feel a lot more social, capable, and alive than I’m used to, but these experiences have also been very interesting.

I want to understand this technology for myself. I want to learn how it can shape the way I make things. I want to learn how it can influence the speed at which I make things. I want to understand how it will change the capabilities of software tools. I want to understand how it might change the world around me.

To do this, I’m going to make something with AI for 365 days. I know “AI” can mean a lot of different things, but when I say “AI”, what I’m referring to is a new growing set of tools and technologies that I have access to as a user of computers and as an amateur computer programmer.

This is not my first experiment with making things for 365 days in a row. In 2017, I made music for 365 days. Prior to this I had not made any music. In 2019, I wrote code for 365 days. Again, prior to this I had not written any code. When I wanted to learn more about UX design, I did the same by making an interface for 365 days.

These 365 day experiments have been transformative for me, allowing me to explore new domains, develop new skills, and form a better understanding of myself and the world around me.

365 days of AI will start today, and here are a few guidelines.

Guidelines

• Each project must have some real world application, no matter how small or niche it may be.
• An output could be of the lowest quality and smallest scope as long as it “works”.
• I do not need to ship before midnight, just before I go to bed.
• In the event I miss a day because of difficult circumstances, I am allowed a “make up” by shipping two things the next day.

Tools and Documentation

• To stick to, maintain a streak, and document this experiment, I’ll be using Futureland, which is a tool that you can use to form habits, learn skills, finish projects, and more. All of my daily creations during 365 days of AI will be documented there.
• At times I will share highlights on Twitter.
• I will use my blog for more refined reflections and documentation, you can subscribe to receive my blog posts via email.

If you are interested in running your own experiments with AI or in any other domain, I encourage you to join me on Futureland so we can progress through them together.