The Drip
Justin brought a BCG finding into this week's conversation: a meaningful share of frontline AI users reported saving at least eight hours a week, while most of that group got little or no guidance about what to do with the time. That gap is more interesting than the efficiency itself. A company can create capacity without creating a better result.
Inside The Bottle
Efficiency is an output. Strategy is a decision about where that output should go.
That distinction matters because AI can make a workflow faster, reduce the work attached to a role, or create room in a budget. None of those outcomes automatically makes a company smarter, more valuable to customers, or better for the people who work there. Capacity is only an option. Leaders still have to choose what happens next.
A useful way to think about that choice has three parts.
Give the capacity a destination. If a company wants faster, more accurate customer quotes, say that. If it wants people spending more time with clients, say that. “Use AI” is not direction. It is a blank at the end of a sentence.
Keep people in the story. Entry-level work often produces future managers and industry leaders. Automating the early tasks without redesigning how people learn can hollow out the talent pipeline. The more interesting question is how AI can accelerate someone's knowledge and ability, then redirect their experience toward work with more ownership and impact.
Make governance enabling, not only restrictive. Risk controls matter, especially around privacy, security, and legal exposure. But a policy that only transfers responsibility to the employee and says “don't screw this up” is not useful governance. Good governance also explains what people can do, where experimentation is valuable, and how their work connects to the company's strategy.
This is why vision comes before guardrails. When people cannot connect AI to a credible company direction, they fill the gap with whatever they hear in their feeds: job loss, abundance, or endless experimentation. A compelling vision gives people something real to aim at.
AI can absolutely make a business leaner. But if leaner was already the need, that need existed before AI. The more durable opportunity is to decide what better looks like, then use the capacity AI creates to move toward it.
Lab Notes
| ■ | Justin's note: Reducing headcount creates budget room. It does not necessarily create a better result. |
| ■ | Kellan's note: The real question is not whether AI saves eight hours. It is who gets the benefit of those eight hours. |
What Stopped Our Scroll
| ■ | BCG: AI at Work 2026 — The report behind our opening conversation: frontline users are saving time, but most get little direction on where that capacity should go. |
| ■ | Gartner: AI layoffs and returns — In its survey of large enterprises using autonomous technologies, workforce cuts were nearly as common among high and low ROI respondents. A smaller payroll is not proof of a better business. |
| ■ | HBR: Redesigning work — Start with how the work should change before deciding which roles should disappear. |
|