Nobody watches robot archery
An archer's job is to be a robot, yet nobody would watch a machine do it perfectly. The test for what to automate: does anyone value watching a human do it?
Thoughts on AI, transformation, leadership, and the craft of getting things done.
An archer's job is to be a robot, yet nobody would watch a machine do it perfectly. The test for what to automate: does anyone value watching a human do it?
When machines take the machine-work, the human-made becomes the premium one. Why the most valuable work of the next decade will carry human fingerprints.
How I built a local-first, encrypted loyalty card wallet for iOS and Android - barcode scanning, on-device OCR, payment card rejection, zero network calls.
How a hand-written scavenger hunt with a muddy stick became Peekabout, the iOS and Android app for parents who want their kids to look up.
When the technical translation barrier collapses, the talent pipeline widens. The organisations that notice first will compound an advantage.
Why everything I build with AI has an expiry date. The artefact is the vehicle; the understanding of your problem space is the asset that compounds.
A teenager downloading freeware in the 1990s wondered who clicked the PayPal coffee links. Twenty-odd years later, I built an app and put one in.
A smart mirror that pulls from eight data sources and uses two AI models to surface what matters before you think to ask - for about 16p a day.
AI is growing total demand for software but shrinking the junior roles that train future seniors. That pipeline gap has a 5-7 year fuse - plan for it now.
Cheaper creative production has always expanded total demand. AI follows the same Jevons pattern - creators who invest in distinctive craft benefit most.