2026: The year of throwing my agency at my health (now with added cyborgism)
A personal AI assistant logs mental health metrics via voice, creating the longest tracking streak ever.
In a detailed personal account, Ruby describes leveraging advanced AI tools to gain unprecedented agency over managing bipolar disorder. After years of failed manual tracking attempts, they built a personalized LLM-assistant system dubbed 'Exobrain.' This system allows for seamless, voice-activated daily logging of mental state metrics, including a custom bipolar index scale ranging from -10 (extreme depression) to +10 (extreme mania). By simply saying 'Hey Exo' and then describing their state, Ruby has maintained a rigorous twice-daily logging streak for three months, capturing valuable longitudinal health data that was previously impossible to collect consistently.
The core innovation is the frictionless integration of AI into daily health routines. The voice recording is sent to a server for processing by DeepSeek's DeepSeek-V3 model, which transcribes and structures the data. This system directly addresses the executive function challenges often exacerbated by bipolar depression, with Ruby estimating a 20% improvement in personal capability. The project represents a practical frontier in personalized health tech, moving beyond generic wellness apps to create a bespoke 'exobrain' that assists with capture, recall, and prioritization, fundamentally changing the user's relationship with their own mental health data and management.
- Built a custom AI 'Exobrain' using voice commands ('Hey Exo'/'Jarvis') for frictionless daily mental state logging.
- Achieved a three-month, twice-daily tracking streak of a custom bipolar index (-10 to +10), the longest ever maintained.
- Estimates the system has made them '20% more of a person' by improving executive function and data capture.
Why It Matters
Demonstrates a practical, user-built AI solution for chronic condition management, moving beyond theory to real-world impact.