Leadership discussion tool · ~35 minutes

The Golden Window

The models are already good enough to change how your organisation works — and right now they're both underpriced and underused. You pay far less than the value you get, and few people around you yet have the focus and skill to capture it. That double gap is the golden window. It won't stay open: usage will be priced to its true worth, and the skill will become ordinary. This session is about what you do while it's open — because the advantage you're really buying isn't the saving, it's the head start.

How to run this — facilitator guide

Timing (35 min)

  • 0–4Frame it: this isn't a pitch to buy AI. It's a decision about timing. The claim on the table is that the return on adopting is unusually high right now and will fall — so waiting has a cost most boards never put a number on.
  • 4–15Thread 1 — the window. Set the two assumptions together, then argue about them. The argument about how long the window stays open is the deliverable.
  • 15–25Thread 2 — capture rate. Score the five dimensions honestly. Low scores are not an indictment; they're the to-do list. Capture the gaps.
  • 25–32Thread 3 — the head start. Move the "when do we start" slider to what the room actually believes, and sit with the gap at the horizon.
  • 32–35Land one commitment. Read the foundation map: which stones got attention, which got none? Write the single move you'll make before the next meeting.

Say this — up front

"Nobody in this room disputes that AI is powerful. The question I want us to answer is narrower and more uncomfortable: is it cheaper and more valuable to start today than it will be in two years — and if so, what is the cost of the delay we're currently choosing by default?"

Say this — before Thread 3

"When this stuff is cheap and everyone has it, the tool stops being the advantage. What's left is everything you built around it — the reworked processes, the fluency, the context only you have, the trust. You can buy the tool later. You can't buy the two years of compounding you skipped."

Note discipline

Every gap the room names gets captured as a foundation stone. If a comment can't be tied to a stone — access, focus & time, skill, verification, workflow, or mandate — it isn't a move yet; ask the speaker to make it one.

Thread 01
The window

Underpriced today. Priced to its worth soon.

Two lines. What you can get from AI keeps rising as the models improve. What you actually pay for that value is being held down right now by a land-grab: fierce competition between labs, subsidised pricing, generous free tiers. The space between them is your surplus — the golden window. As usage gets metered and priced to what it's worth, the lines converge. Set the two assumptions and watch it close.

Value you can get What you pay Your surplus

Illustrative model — the shape of the argument, not a forecast. "What you pay" is the effective price of getting real work done, not the sticker price per token: per-token rates may keep falling while what a finished outcome costs you rises, as metering tightens, free tiers shrink, and the cheapest serious work stops being free.

How subsidised is pricing today? Heavy
How far below its value is AI priced right now?
Near its worthHeavily subsidised
Years until it's priced to its worth 3 yr
How long before the price catches up to the value?
FastSlow

What the room is assuming

Thread 02
Capture rate

An open window you can't reach through is worth nothing.

Here's the catch the price line hides: the value takes focus and technical fluency to extract, so today most of it goes uncaptured. Score your organisation honestly on the five things it takes to reach through the window. A weak link anywhere caps the whole — you can't verify your way past no access, or prompt your way past no time to learn.

Your capture rate

0% of the surplus the window is offering

Thread 03
The head start

The tool becomes a commodity. The foundation doesn't.

Picture two organisations in the same rising-AI world. One starts building capability now; the other waits until it's cheap and everyone has it. When the waiter finally starts, they buy the same ubiquitous tools — but they're climbing a learning curve the first mover is already near the top of. What compounds isn't the tool. It's the reworked workflows, the fluency, the context only you hold, the earned trust. Move the slider to what this room actually believes.

Start now Wait, then buy the same tools The gap you can't buy back

Illustrative model. Both organisations end up with identical tools — the gap is everything built around them, which accrues with time and can't be purchased in a lump.

When does the one who waits start? 2 yr from now
Later feels safer — cheaper tools, fewer mistakes, someone else's lessons. This is what it costs.
Start in 1 yrStart in 5 yr

The cost of waiting

Not the last word on this — The Efficiency Dividend takes the money side into the boardroom, and The Expectation Treadmill shows why standing still is its own kind of moving.