Don Scheibenreif’s diagnosis is uncomfortable because it is not about the technology: adoption is running bottom-up while denial runs top-down. Over ninety percent of CEOs use AI, yet nobody on the executive team scores as AI savvy—and the chief HR officer ranks last, right as the work becomes redesigning the systems employees engage with. Meanwhile the disintermediation Gartner predicted in 2014 is arriving properly this time, because customers want easy and friction is what intermediaries sell.
These AI assists don’t work for us. They work for the people that made them.
—Don ScheibenreifDon Scheibenreif in 2023 against Don Scheibenreif in 2026—what changed about machine customers, what held, and the question three years did not answer.
Robb Wilson on why record search and social ad revenue is a lagging indicator—and why the queries that used to carry the money are moving into agents.
The introduction to Matthew Barzun’s The Power of Giving Away Power—why pretending is the price we pay to avoid uncertainty, and what happens when a campaign gives the voter file away.
Robb Wilson compares LangGraph, CrewAI, and OneReach GSX by asking what happens when an agent makes a wrong call—and which stack’s defaults already believe what your auditor believes.
Former NASA CKO Roger Forsgren points to ASRS—the voluntary reporting system behind commercial aviation’s safety record.
Alla Slesarenko on the asymmetry between building agents and governing them—and why accountability outlasts the project plan.
Mathieu Penot on why AI can already teach lessons but cannot replace teachers—and why keeping humans in guiding roles is a design constraint.
Rich Weborg on why AI makes design artifacts cheaper—and why taste, coherence, and evaluation become the designer’s real leverage.
Good Will Hunting, canonically correct knowledge, and why LLMs regurgitate books but cannot tell you what your Sistine Chapel smells like.