Afterthought

"There are these two young fish swimming along and they happen to meet an older fish swimming the other way, who nods at them and says “Morning, boys. How’s the water?” And the two young fish swim on for a bit, and then eventually one of them looks over at the other and goes “What the hell is water?”" -David Foster Wallace, This is Water

Often, the most important things about our lives are the ones that become hidden in plain sight. Like water to a fish, our perceptions of ourselves, our loved ones, and the world around us are often steeped in intentions we have trouble recognizing.

We want to change how journaling works.

We want to help users reflect by applying LLM AI technology to their journaling. Afterthought picks up on your intention (whether you're speaking about a goal, an experience, a relationship, or something coming up) as well as your mood in order to give you a benchmark to better understand yourself. Pairing this analysis with a summary of recent journals and reflection prompts, we intend to make journaling go deeper. Not only does Afterthought inform you of patterns, it actively prompts you to question the things you've been telling yourself.

How we built it

We started out in the cohere playground, brainstorming what kind of categories would be valuable. We went into our own notes apps and journals to see how we might use the app. From there, we began constructing a web app, a categorize script, and a summarize script. We later married these elements together.

Our dataset was first formed by coming up with out own examples of a variety of possible notes. Then, we asked chatGPT to come up with a variety of similar notes, varying the way these sentences were structured and the names mentioned. We pruned this list to make sure it was a balanced representation of possible notes.

Challenges we experienced

Nobody on our team had used all of these technologies before, meaning that there was something we all learned. In addition to the typical hackathon blues (drowsiness and sore tummies), we spent a lot of time getting around learning curves.

What's next

We'd like to add features that allow for people to track specific relationships/concepts, rather than these general categories. Additionally, creating a way to classify and reflect upon ideas/art is a direction we'd be interested in exploring.