Present a model this lean carelessly and it reads as "humans get replaced" - wrong, and fatal to adoption. The honest picture is expansion, with a real transition to manage. Getting it right matters: the scary version isn't just demoralising, it's less accurate.
01 - The fallacy
"AI does the work, so there are fewer jobs" assumes a fixed amount of work to go around. Historically that's been wrong.
When the cost of producing something falls, demand for it expands - often faster than the efficiency gain. ATMs didn't end bank tellers; cheaper branches meant more branches. Spreadsheets didn't end accountants; they expanded what analysis could do. Compilers didn't end programmers; they exploded the amount of software.
Cheaper building doesn't mean less building. It usually means far more - much of it never worth doing before. Companies attempt more; the pie grows.
Make something cheaper and the usual result isn't less of it. It's an explosion of it - Jevons, every time.
02 - The shape of the work
The codifiable-middle roles shrink; edge work and proxy-orchestration grow.
Your day shifts toward two things: creating new edge, and fielding the questions your proxies escalate - yours and other people's. Your "team" becomes a web of digital representations that handle their own interior and promote their boundary to you.
The work moves up the value stack - toward creativity, judgment, supervision. Less drudgery, more leverage.
You sit at the hub. The proxies handle the interior; they escalate only the boundary - and the boundary is where your judgment lives.
03 - Kept honest
"Don't scare people" must mean telling the true hopeful version, not pretending there's nothing to manage.
Aggregate optimism and individual disruption are both true. "More work in total, eventually" doesn't mean no one is displaced. Middle roles shrink, and the people in them have to move to edge work - navigable, but not automatic, and deserving deliberate reskilling, not a shrug.
Distribution is a genuine open question. A growing pie doesn't guarantee the gains reach edge-workers rather than concentrating. That depends on choices, not physics, and it's contested.
Per-person expansion is bounded by run-capacity. Proxies amplify you, but only up to what one human plus their proxies can sustainably orchestrate. Real expansion, not infinite.
Aggregate optimism and individual disruption are both true. The honest version says so.
04 - The framing that lands
Said well, this isn't a threat.
The proxies take the commodity off your plate so your scarce judgment goes to the edge that differentiates the company - and because the company can now attempt more, there's more edge to create, not less. People are amplified and pointed at the part of the work that was always theirs.
That's the version to lead with: it's both more accurate and the only one that gets adopted.
People are amplified and pointed at the part of the work that was always theirs.
Cheaper building doesn't mean less building - it means far more, and the displaced middle re-emerges higher up. The shape of the work shifts toward creating edge and orchestrating a web of proxies that escalate their boundaries to you. Be honest that the transition disrupts individuals and that distribution is contested - then lead with the true hopeful version, because it's both more accurate and the only one that gets adopted.