Speaker engaging audience during a professional development event in Australia.

Which Act of AI Are You Actually In?

29 Oct 2026 | By Phuel

Your whole organisation might be using AI every day. That doesn’t mean you’ve got anywhere with it, and according to this episode, most teams are stuck somewhere between usage and value without realising it.

Host Nathy Gaffney splits the conversation into three acts: adoption, replacement and reimagination. Andrew Webster, VP of Organisational Innovation at ExperiencePoint, calls in from Toronto with data from the Americas, Europe and China. Jake James, Phuel’s senior facilitator and AI practice lead, brings what he sees in the room with leadership teams each week. Andrew starts with a recent trip to Saudi Arabia, where adoption is often tracked by logins and prompts per day. Fine early on, he says. But a team that writes a mountain of emails with AI will score high on usage without creating much of anything. Jake disagrees that organisations overestimate their progress. He reckons they just thought adoption was the finish line. He’s had teams with near-total uptake light up in workshops over basic features they’d never tried.

Being stuck in act one looks a lot like being busier than ever and a bit let down. Andrew calls it AI underwhelm. Jake is kinder about ‘work slop’. Calling it slop, he says, is like telling his three-year-old the drawings he brings home from daycare are rubbish. It’s a messy stage. Things often go backwards before they go forwards. Nathy points out the toll on people who are used to being the expert and now have to be a beginner over and over.

Who has it worse, the people who don’t want AI or the people who aren’t allowed it? Jake says the second group. Curious people under pressure will go looking for tools that work, and that’s where the risk creeps in. Andrew has two examples. One is a global pharma company that tells the world AI is core to its business but won’t give its salespeople licences. The other is a Fortune 300 CEO who warned staff to fear colleagues who know more about AI, then gave them no access to learn it.

Andrew’s line that optimisation is where organisations go to die is probably the one that sticks. He toured a Chinese manufacturer that’s the best in the world at a very specific component. Demand is dropping. And because it has replaced so many people, there’s no one left to go and find out what customers need next. Indeed is the opposite case: it wins back time with AI and moves those people onto bigger problems. Jake adds his ‘hugging the cactus’ exercise, where workshop participants ask AI how their role will change over the next five years. One self-described data lover came back rattled. AI would soon do her favourite part of the job better than her, and she didn’t want to spend her days checking its work.

Then a small admission from Andrew. He has never had AI write an email, mostly because he likes writing and wants people to know he made the effort. He mentions Jeffrey Sanchez-Burks’ idea of an imperfect human premium: a typo in an email is starting to read as a sign of care. Pipedrive’s new AI coach Nova comes up as a live example of reimagination done properly.

To get unstuck, Andrew says celebrate how much you’ve already changed and use that momentum. Jake wants leaders to share their own messy prompts and mistakes so their teams feel safe to try. Andrew also warns about agentic sprawl, which is what happens when people build AI apps nobody needed. His fix is to fall in love with the problem before the solution.

This episode is worth your time if your team ticks every adoption box but can’t point to what’s actually changed, not if you’re after another list of AI tools to try.

Ready to amplify your impact?

Australian professionals at a leadership and team performance workshop.