Use AI for work
Concrete task-based playbooks — email, meetings, research, weekly reports, second-brain capture.
Identify what to automate: the weekly audit
Track time for a week, find the 5 most mechanical tasks, decide which actually belong to AI. The trap of automating things that should not exist at all.
The email triage playbook
Auto-categorize, draft replies, flag VIPs, archive newsletters. Concrete setup using Lindy, Zapier, or Claude.
Replace your weekly report with AI
Pipe Slack, calendar, and notes into AI to generate weekly status, project updates, or board summaries.
Calendar and meeting prep: never walk in cold again
Reclaim and Motion for the calendar, Granola and Crystal for the prep. A 5-minute morning ritual that buys you the whole day.
From voice notes to a second brain
Capture thoughts as voice on the move, transcribe with Granola or Otter, route to Notion, Obsidian, or Mem with auto-tagging. Stop losing ideas.
How to use AI for research
A 4-step model — map, find gaps, pressure-test, turn into output. With the failure modes (hallucinated sources, smeared comparisons, confident guesses) and how to verify before trusting.
How to write better emails with AI
A 5-input model that stops AI emails from sounding like LinkedIn spam. Plus revision-mode (the underused move), a ban-list, and when not to use AI for email at all.
How to use AI for meetings, notes, and follow-ups
Capture is table stakes. The value is decision extraction, action items with owners, and same-hour follow-ups. A practical model and the prompts that turn transcripts into work.
How to use AI for spreadsheets and analysis
AI as analyst assistant — formula generation, cleanup, segmentation, anomaly spotting. With the dangerous failure modes (fabricated totals, spurious precision, causal claims from correlation) and when to use code-interpreter mode instead.
Build with AI
Ready to build software with AI? Start with vibe coding and graduate from there.