I’m revisiting some of the articles I read in 2019 (5 years ago, wow!). One of them is Aaron Z Lewis’ wonderful essay on ‘Designing a Dynamic Journal’. There are some fantastic excerpts on the medium of journalling, and the broader possibilties for this tool for interacting with oneself. Examples:

Life is an RPG — What if your journal could interpret what you’re writing and help you conceptualize various parts of your daily journey in the form of a video game? I know this isn’t what Aaron is saying here, but still:

“My journaling habit started in 4th grade, when my school librarian gave me a fat notebook with a squishy tie-dye cover. Every day, I wrote something about my extraordinarily complex 10-year-old life. Mostly, I documented my video game successes and failures—new levels and weapons unlocked, defeats at the hand of my arch-enemies, etc. Almost everything about my life has changed since then, but I’ve never stopped journaling.”

Today, with LLMs increasingly becoming capable, a lot of the suggestions feel far more in reach. Both in terms of LLMs enabling the feature and code copilot’s making it cheaper to deploy the code for non-LLM features.

On the Shift in Capacity for Cognitive Manipulation by introducing a ‘Delete Button’

“As a digital native, it’s hard for me to imagine writing without the convenience (crutch?) of the backspace key and copy/paste. But to Engelbart and his contemporaries, the word processor was mind-blowing. His manifesto made me realize that handwritten thoughts are different from typewriter thoughts are different from Microsoft Word thoughts are different from Notion thoughts. I’ve obviously written things on paper before, but always with the assumption that they’ll graduate to the digital world where I can easily mold and manipulate them. The word processor gave us the freedom to juggle the record of our thoughts, which has in turn shaped the way we think.”

Fun fact: Neil Postman writes his first drafts on pen and paper specifically BECAUSE this ‘deletability’ is too powerful an affordance. Neurotic editing tends to take over instead of writing with the rythym of slower cognition and pen/paper.

LLMs that can predict your next thought:

“Predictive text. A model trained on your thought collection could try to predict what you’re going to type next. Typing in your dynamic journal could feel like brainstorming with your past selves.

I’d be interested in the ways you can configure this to predict what a ‘version’ of yourself might say. For example, ‘Ecstatic Me’ or ‘2019 Me’. We contain multitudes, after all.

Finally, on the Representation of Self Knowledge (The Journal as a Mirror):

“A dynamic journal is just one tool of many that people might use for self-reflection. It doesn’t provide any answers or prescriptions, but it might help you ask some interesting questions. The invention of the data graphic in the 1700s revolutionized the way we understand complex systems. The invention of cartography changed the way we understand our place in the world. Representations matter because they shape perception. Dynamic journaling is a way of representing thoughts that might help us understand ourselves just a little bit better. If nothing else, it makes for an interesting thought experiment.”

On this idea of ‘digital representations of self, I’m reminded of a girl that was on the tube, on her way to Heathrow (like myself at the time). She was going through pictures of herself from during her holiday. She must have looked at the same picture for at least 30 minutes.. God bless her.

---

Exciting possibilities. Given the choice, I would buy existing distribution (ie Day One Journal), and try and impose these experiments on that population, for the sheer joy of seeing new forms of cognitive technology come to life.