Three layers of AI, and the one that got me home
Three days at Dublin Airport showed three layers of AI: optimisation, judgement and everyday automation. Most of the return sits in the least glamorous.
The problem is not that AI does not work. It is that the ruler is wrong.
Boards keep quoting MIT’s finding that 95 per cent of enterprise AI projects show no return. Why that reading is wrong, and what it may cost to wait.