1️⃣ HR has to go first McKinsey says HR now has a dual AI mandate: redesign work across the enterprise and rebuild its own function at the same time. It estimates current technology could automate activities equal to 57% of US work hours.
💡 Why it matters Workforce planning cannot stay a headcount exercise. Leaders need task maps, capability plans and clear rules for where agents enter the operating model.
☕ Coffee talk Can HR credibly tell the business to redesign work if its own processes still look untouched?
2️⃣ Innovation needs an owner after the demo INSEAD argues open innovation often stalls in the proof-of-concept death valley: pilots, accelerators and partnerships launch, but promising ideas fail to move into operations.
💡 Why it matters Experimentation is the easy part. Scaling needs a business owner, integration path, data fit and governance before the first pilot starts.
☕ Coffee talk Who owns the handoff after the startup demo, or is everyone just admiring the sandbox?
3️⃣ AI makes performance harder to judge IBM’s new global study says 93% of executives find AI-enabled work harder to evaluate, while only 43% of employees report employer-provided AI training.
💡 Why it matters Managers are being asked to judge output they did not see produced. Performance systems need new standards for judgment, disclosure and challenge.
☕ Coffee talk Who is ready to explain a promotion decision when half the work came through an AI workflow?