Signing in

You will be sent to MillerKnoll sign-in.

AI

The footprint, and using AI on purpose

AI has a real cost in energy and water. Concern is warranted. Panic is not. Both of those are true at once, and holding both is the grown-up position.

Lesson 5

Here is the part most people get wrong. It is not training the models that drives the footprint. It is everyday use, an estimated eight to nine of every ten units of energy. Which means the footprint is in our hands, not just in a data center somewhere.

We measured our own. And the lever you hold turns out to be a simple one: use it on purpose. Get the prompt right the first time instead of re-asking it ten times. That is a smaller footprint and a faster answer at once.

Reflect

Same task, summarize last week's team metrics for your manager. Two ways to ask. One takes several runs; the other takes one.

Vague prompt

Summarize the metrics.

    Each run uses energy and time, and you still do not have the answer yet.

    Specific prompt

    Write a 150-word summary of last week's team metrics for my manager. Lead with the one trend that matters most.

      One run. Same task, answer on the first try.

      Rough order-of-magnitude, not a meter. Daily use and re-asking multiply cost. This is why Lesson 3 asked you to be specific.

      Check yourself

      Where does most of AI's energy footprint actually come from?

      Before you go

      One habit that shrinks your footprint and saves you time: get the prompt right the first time.

      Did you run this in Copilot? Mark complete when you have tried it.

      Next lesson: Where you go from here →