stack smarter ways of work
honest takes on AI and work. research-backed opinions on what's actually changing, what the data shows, and how humans and AI work best together.
Your 'Second Brain' Is Write-Only. Here's Why Most Notes Never Get Revisited.
Most personal knowledge management systems fail not because of bad tools, but because capturing notes and revisiting them are two completely different habits — and we only ever build the first one.
Your Designers and PMs Can Ship Code Now. Here's What That Actually Means.
Designers and PMs are shipping real code using AI tools. Here's what that shift means for team structure, code ownership, and who actually controls the bottleneck.
Garry Tan's gstack: Why a 23-Step Workflow Beat 'Just Prompt the AI'
Garry Tan's gstack breaks AI-assisted coding into 23 specialist roles, and understanding why that structure matters could change how your whole team works with AI.
29% of Fortune 500 Companies Are Paying AI Startups: Here's Which Ones and What They're Buying
A data teardown of enterprise AI adoption reveals that only 29% of Fortune 500 companies have live, paying contracts with AI startups — and why the real story is in the 71% that didn't convert.
Productized Services Over Hourly Rates: The Pricing Model That Survives AI Undercutting
Learn why productized services with fixed-scope, fixed-price packaging protect your freelance income as AI drives hourly rates toward zero.
The 47-Second Attention Crisis: What the Research Actually Says (and Doesn't)
The 47-second attention span statistic is real but widely misread — the actual crisis is fragmented attention caused by tool design, not shrinking biological capacity.
The 29% Trust Problem: Why Your Team Won't Actually Use AI Tools (Even When You Mandate It)
Developer trust in AI code tools has dropped to 29% even as usage hits 84% — here's why mandating AI adoption backfires and what managers should do instead.
The White-Collar Recession That Hasn't Happened Yet: Why Anthropic's Own Data Shows a 94%-to-33% Gap
Anthropic's 2026 labour market report finds AI could theoretically handle 94% of computer and maths tasks, but professionals only use it for 33% — here's what that gap actually means for white-collar workers and what to do about it.
Why Your Freelance Rates Just Shifted: The Economics of Independent Work in 2026
Freelance pricing pressure is rising in 2026 as more workers go independent, but solos who specialise in high-touch, revenue-driving work can still command stronger rates than ever.
The Rule of Two: Why Two Humans + AI Beats Fifteen People (And Why This Matters for Your Team)
Emerging research shows two humans working with AI can match or beat much larger teams — here's what that means for how you hire, structure, and manage work.
Deskilling Is Real, But It's Not Inevitable: How to Use AI Without Losing Your Edge
AI deskilling is real and measurable, but the teams building skills faster with AI share one trait: they treat it as a thinking partner, not an answer machine.
How Markdown Files Became the Control Layer for AI Tools
A practical guide to AI markdown configuration: how CLAUDE.md, cursor rules, and .github/copilot-instructions.md files work, what the research actually shows about their value, and how to write them so they help rather than hurt.
Meeting Summaries That Actually Stick: Why Auto-Transcription Alone Fails
Auto-transcription captures words but misses the point — here's how to turn meeting notes into action items, owners, and closed loops with your project tools.
84% of Engineers Report AI Productivity Gains — Here's What They Have in Common
84% of engineers report AI productivity gains, but the data shows they're gaining time in verification and review, not code writing — here's what separates the engineers seeing real results from those just feeling busy.
Custom AI Agents vs. ChatGPT: When to Build vs. When to Subscribe
A practical decision framework for choosing between a ChatGPT subscription and a custom AI agent, with real numbers on cost, ROI, and when each approach actually makes sense.
The Competitive Intelligence Agent You Can Build in an Hour (And Why Your Competitor Hasn't Yet)
Learn how to build a competitive intelligence AI agent in under an hour that automatically pulls news, job postings, and competitor signals into Slack briefings, with a concrete step-by-step walkthrough.
McKinsey's 1.5M Hour Gain: How to Find Your Organization's Biggest AI Opportunity
McKinsey's internal AI pilot saved a large bank 1.5 million hours annually — here's the framework to find your organisation's equivalent bottleneck and size the real ROI before you commit.
The 60% AI Productivity Gain Is Real. What the Teams Achieving It Have in Common
Most teams quoting the 60% AI productivity gain never achieve it. Here's what the teams that do actually have in common, and what everyone else is getting wrong.