What I'm using AI for

When speaking to friends who aren’t using AI tools much and I begin to evangelise, I often get asked “what do you actually use it for?”. So here’s a high-level summary. Note, it’s not exhaustive but aims to give an indication of the breadth of use cases.

Also, the examples shown are just examples, not a complete list.

Updated Jan 2026

• Codegen – writing a lot of code (recent examples: Flow, Kotlin app for my touchscreen device without knowing any Kotlin)
• Code review – reviewing code for quality, security, architecture, particularly when a new model comes out!
• Brainstorming – product and implementation ideas for open source projects
• Search – particularly when my search subject isn’t a specific thing (e.g. “what’s the essay Paul Graham wrote about gaining knowledge from reading books but not knowing which book it came from”)
• Troubleshooting – solving issues (e.g. Spotify Connect on home network)
• Internal tooling – creating transient tools (e.g. Github Spark tool for DX onboarding data collection)
• Hardware/embedded software and config – Adding a Home Assistant bluetooth proxy to an ESP32 device, and turning off the LED
• Document outlining – creating outlines for business cases, checking I’ve not missed key points
• Articulation help – constructing arguments when my head’s cloudy
• Codebase summarisation – understanding new codebases quickly
• Home manual grounded agent – querying appliance manuals (“what does red flashing light mean on my dishwasher?”, “what is error code E123?”, “when do I need to clean my tumble dryer filter?”)
• System administration – managing my machine and server (e.g. ffmpeg commands via Claude Code)
• Log analysis – grokking logs quickly, to both find things and understand them
• Weekly summaries – reviewing Teams messages, emails, docs, and notes changes
• Image generation – creating images for presentations, documents (Engineering Chapter brand)
• Learning plans – structured learning (e.g. “create me a 3 day learning plan, 1 hour a day, to learn evals”)
• Legal document parsing – helped loads with house purchase
• Product research – understanding specs (e.g. speakers)
• Data reformatting – converting images to CSV, paragraphs to bullet points
• Bringing old photos to life – recently colourised a load of my mum’s wedding photos
• Making videos – Descript
• Platform administration – see lf