Elev-8 Performance

AI is having its moment. And to be fair, it is amazing at the stuff that slows people down - ideation, first drafts, automation, turning a messy thought into something usable. It’s basically a shortcut through the blank page.

But here’s what’s becoming obvious:

AI still isn’t a one-stop shop. Most “real work” is a loop ... AI does a chunk, a human makes a judgement call, another tool gets involved, context gets layered in, then someone sanity-checks it against reality.

That’s why the organisations getting value aren’t the ones with the most licences. They’re the ones with the strongest human-in-the-loop capability.

Because the biggest blockers aren’t necessarily technical. They’re likely to be human.

People aren’t just thinking “how do I use it?”
They’re thinking: Is this cheating? Am I allowed? What if it’s wrong? Will I look lazy? Who’s accountable if it goes sideways?
That’s where leadership comes in.

If leaders don’t activate the climate for AI (permission, clarity, role-modelling, and a grown-up approach to risk), people either avoid it entirely… or use it quietly and inconsistently, which is how quality problems creep in.

Thing to try: “The Human-in-the-Loop focus” (10 mins, in any team meeting)

Agree this out loud as a team:

“We can use AI for speed… but humans own the judgement.”

Then set three simple rules in plain English:

1. where AI is encouraged (ideation, drafting, summarising, admin work)
2. where humans must take the wheel (decisions, customer impact, risk, sign-off)
3. what “good use” looks like (say when you’ve used it, sanity-check, don’t blindly paste copy, etc)

It sounds small, but it does something massive... it removes the awkward uncertainty that stops adoption and it protects quality!

In other words: we don’t need humans out of the loop. We need humans in charge of the loop.

If you want AI to lift performance (not just output volume), stop obsessing over tools and start building the climate that makes smart adoption normal.

Most management development doesn’t fail because the content is “bad”.

It fails because it creates 'knowing', not 'doing'.

In Dropping the C-Bomb, we’re pretty blunt about the fix... Great leaders “forget abstract concepts and theories” and instead create clarity about what good looks like, then help people transfer it into real conversations.

And that only happens when coaching becomes part of the team climate (the day-to-day cues people experience in the “here and now”.)

Try this: Coaching in-the-moment (the 90-second version):

We call this coaching in-the-flow-of-work because it’s faster, more human, and it actually sticks.

Next time you spot a wobble, use COJ: Coaching Judgement:

-“I noticed you chose to do X…” (fact)
-“What were you aiming to achieve?” (intent)
-“How appropriate was the result you got?” (outcome)
-“What options did you consider?” (thinking)
-“Next time, what would you do differently?” (new thinking)

That’s straight from the book, because it does something most “feedback models” don’t... it coaches the thought process, not just the behaviour.

Why this matters?

Old-school monthly coaching is too infrequent (and leaders can’t recall enough detail anyway). The recency bias is real - coach closer to the moment, and capability builds faster!

If you’ve got managers who “know the model” but avoid the moment, don’t add more learning.

Add more in-the-moment coaching. It’s the cheapest performance lever you’ve got.

PS - not got a copy of the book yet? Grab one here - https://amzn.to/4bJtwWq?utm_source=facebook&utm_medium=Social%20post&utm_campaign=BAU