For awhile now I've had in mind a locally ran AI ADHD assistant. Using always on ASR for interactions (removing friction) it would keep track of todos, help refocus from distractions, and work as a personal knowledge base to aid with memory.
The key to this would be everything running locally, privacy preserving. The LLM would observe everything that is happening on screen, notice when someone is distracted by reddit (or HN!) and help refocus people.
Perfectly doable on even a moderately spec'd MacBook now days.
The key to this would be everything running locally, privacy preserving. The LLM would observe everything that is happening on screen, notice when someone is distracted by reddit (or HN!) and help refocus people.
Perfectly doable on even a moderately spec'd MacBook now days.