[Incremental System] IRL Status Screen

Complete game-like status screen, with experience and level tracking, in real life. * bullet-point voice-transcribed journal entries -> actions, skills, and characterizations -> statuses. *

Comment

Overview:

Introduction

Hook: https://www.reddit.com/r/funny/comments/19ej6st/a_good_experience/

The vision:

  • How can we motivate ourselves? How can we know exactly who we are?
  • The solution to both of these questions is a single-source-of-truth quantifiable ‘status’. Such a status would display a state of ourselves that can be understood in quantifiable data (knowledge of who we are), and changes to it as we act would serve as direct immediate positive feedback (motivation).
  • In games and media, this is commonly seen as a “status screen”-- a screen of the main character / “players” that displays the users’ level, and statistics like ‘hp / mp’, ‘int’, ‘str’, etc.
  • If such a screen existed – an extremely convenient way to see our own status and statistics, updated automatically, when wouldn’t that be amazing?

Implementation details:

The foundation of such a status screen, implemented into real life, would contain 2 things: Experience (exp), and levels (lv). Experience: a measure of time Levels: a measure of skill Each of these can be implemented much more straightforwardly using LLMs compared to standard hard-coding. Cleverly thinking, these processes would look like: Information input for inference: an easy voice-to-text module that allows the user to rapidly and conveniently input information. These will be called “journal entries” or “entries”. CDAG: characteristic directed acyclic graph; characteristic quantity set graph A DAG that represents a graph of human characteristics: In its ideal form, it would represent (parent to child = more general to more specific): Characteristics (intelligence, wisdom; fitness, endurance; productivity; …’) -> skills (memorization; bench press technique; habit forming) -> actions (studying; exercising; scheduling) Edge weights represents proportions. Experience system: Pipeline: User entry -> “actions” the user took and the “time” it took -> convert into exp gain (eg: 30 min = 1 exp) -> ‘forward feed’ through CDAG smartly and recursively -> generate total exp changes of all nodes -> return results back to user and store results. Inspired by feedforward and backpropagation algorithms in feed-forward-neural-networks Level system: rough experience approximation

My process:

gemini AI studio iterations to solidify idea -> write a google doc on project direction precision and goals -> migrate to VScode editor using pure indexDB backend -> migrate to firebase backend for primary source of truth with full authentication handling. Each migration was a source of trouble and required much debugging. However ai-agents smoothened things out a great deal. The AI-agent basically acted another (or even 2, 3) team member!

Final product:

A complete integrated (pipelined) modularized voice to status screen system with stable backend storage per user and user authentication (with anonymous guest handling).

the code repo:

https://github.com/richardhcli/self-statistics-system-foundation (README and documentation is not updated - ran out of time! Please look at branches ("origin/documentationUpdate") for documentation updates).

More information and details (per entire repo, per levels, per features) found on github repository: https://github.com/richardhcli/self-statistics-system-foundation

Built With

Share this project:

Updates

posted an update —

Known Bugs (oops! I didn't encounter this in bug testing in my app / I did not fix this in time):

  • Multiple tabs open fails authentication: if you log in on one tab, open another tab, and then log in, authentication fails. Please do not do this when testing my firebase app! You should never have to do this

Log in or sign up for Devpost to join the conversation.