Updates on AI in the physical economy.
Tips for soon-to-be AI-Native companies.
90% of industrial products companies are implementing an AI strategy. 20% have it embedded. The other 70 points are sitting in sales and marketing.
Construction, trucking and manufacturing sit at the bottom of every AI adoption measure. Three methods, one answer, and it costs more than anyone counts.
94.5% of fleets bill detention. Fewer than half of those invoices get paid. On a 75-truck fleet that gap is roughly $308,000 a year, lost to paperwork.
17.2 RFIs per $1M of construction cost. Eight hours each. A 9.7 day median reply. What bid week and RFI turnaround actually cost a contractor.
Two thirds of staffing firms are "doing AI." One in seven has bought anything, and almost all of it stopped at screening. The hours are somewhere else.
A federal audit found contractors had not completed 69% of sampled work orders, with sign-offs on work visibly undone. At a 3.6% margin, that arithmetic is brutal.
A proof of concept is TRL-3. A pilot is TRL-7, running in an operational environment. The US government codified the difference decades ago. Most AI vendors ignore it.
57% of employees hide their AI use. Your usage dashboard is measuring the wrong thing. What a rollout scorecard measures instead, department by department.
89% of executives believe they could switch AI vendors within a month. Of those who tried, 58% hit failure or unexpected effort. The lock-in is in the rules, not the API.