This session is not about opening a tool. It is about positioning ourselves before the practical work begins. Where AI actually is right now, where each of us sits in the cohort, what we are committing to across the programme, and the principle that holds it all together. People lead. AI follows.
- Locate yourself on the Talkers / Users / Shapers continuum.
- Understand the program's promise, the six-session arc, and what it asks of you.
- Make a first pass at your AI Role Plan, picking the area of your role this program will work on with you.
- Leave with concrete work between sessions grounded in your real work, and the journal habit started.
Session Slides
34 slidesFor the best view, open the slides in a new tab. Some content may be clipped when viewed in this embedded panel.
Session Recording
1h 51m · Recorded 6 May 2026Recording is for cohort use only. Please do not share the link outside the programme.
Positioning AI (summary)
A look back at Session 1A look back at where we began on 6 May. The ideas we covered together, what came up in our conversation, the prompting techniques worth taking with you, and the actions to take across the fortnight ahead. Click any section to open it.
Eight ideas formed the backbone of Session 1, and you will see all of them return across the rest of the programme. They are worth a second read before you start experimenting in your own work.
What AI actually is
We pulled the words artificial and intelligence apart and landed on a working definition. A human-made ability to perceive, discern and choose. Because it was built by people for people, our work and our judgement matter more, not less.
Where we are on the curve
The technology is well past the tipping point but adoption inside organisations is still catching up. The gap between the two is what creates the swirl most workplaces are feeling right now.
OpenAI's five stages
The roadmap moves from chatbots to reasoners, then to agents, innovators and eventually whole organisations, with each stage building on the last. The chatbots came first on purpose, designed to meet our human needs for certainty, harmony, recognition, momentum and relief, which is why they took off the way they did.
Talkers, users, shapers
If you walk into a room of a hundred people, roughly seventy are still talking about AI, about twenty five are using it like an enhanced Google, and only five are shaping it. The real value sits in the shaping, which is where this programme is taking you.
People Lead. AI Follows.
This is the philosophy that holds the whole programme together. AI is not here to turn us into robots or to take the lead. It is something we shape and direct, with people staying firmly in front of the work.
Top down meets bottom up
Humans tend to plan from the top down, moving from strategy to objectives to projects to tasks. AI works the other way around, starting from tasks and gradually building up. Knowing the two work in opposite directions helps you decide where to bring AI in and where to keep your own judgement out in front.
The silo cascade
Divisions used to silo from each other. Remote work then started siloing teams within those divisions. Now AI brings a new risk, where individuals can end up siloed because they can run whole workflows alone. The real uplift sits in collaboration, not in going it alone.
Your Individual Intelligence System
The focus of this programme is your own role and the work you have authority over. That is where we will build the agents and workflows you can actually use, before extending the same thinking across teams and the wider organisation.
Some of the most useful parts of Session 1 came from the conversation rather than the content. These were a few of the threads that opened up across the room.
These are the prompting techniques that surfaced across Session 1. They are not meant to be used all at once. Try one or two across the fortnight, notice the difference they make in the work, and let the rest sit until you are ready for them.
Tell the AI who the audience is, what time span you are working with, what shape of output you want and what rules you would like it to follow. Most of the difference between a generic response and a useful one comes from how much context you provide upfront.
If you want the response to land in a particular shape or form, send it a document or sample as a reference. Often it will answer most of the questions you would have asked before you even get to them.
Build a rule into your prompt that asks the AI to pause and check what it needs from you before producing anything. The slight slowdown almost always produces a stronger and more relevant result.
Be specific about whether you want the response to sound personal, formal, technical or somewhere in between. The AI adapts well when you ask, but it will default to corporate and generic if you do not.
When the AI comes back with a list of questions for you, ask it to answer them itself based on what it already knows about your organisation. You then move into the role of editor rather than information gatherer, which is a much better use of your time.
When the AI asks you a question and you do not know the answer, ask it to give you a few options to choose from. Picking A, B or C is much faster than trying to formulate the perfect answer from scratch.
If you are working on something substantial, ask the AI to go through each section one at a time rather than producing the whole thing at once. It stops you drowning in forty questions and lets you make decisions in smaller, more manageable chunks.
It is rare for the first response to be the best one. Ask the AI to refine what it has given you, whether that means making it more concise or going deeper into the detail. The second pass is almost always sharper and more useful than the first.
If something surprises you or shows up in a way you did not expect, ask the AI how it got there. Where did that information come from, how does it work, and what was it drawing on. The AI is happy to explain itself, and the answers are often more instructive than the original response.
When you hit a moment of frustration or get stuck, stop and tell the AI exactly that. Describe what is in your way and ask how it could help you move through it. Most of us try to push through obstacles on our own, but with AI it works far better to name them out loud.
If a response feels off or too smooth, ask the AI to defend the claim. Is that actually true. Where is the evidence. Both generic outputs and confident hallucinations tend to fall apart when you ask a direct question like that.
Ask the AI to critique its own response before showing it to you, ideally against a clear set of criteria. It produces a noticeably stronger answer when it has had to evaluate its own work first.
The biggest difference between people who get value from AI and people who do not is whether they bring real work to it. Abstract examples and made up scenarios will not give you the kind of insight that comes from working on something that actually matters in your role.
Resist the temptation to keep refining until the AI gives you something perfect. Get it to about eighty percent and then bring your own judgement and editing to the last twenty. That is where you add the most value to the work.
Four things to take into the next two weeks. None of them should take very long, and they build on each other. Start small, experiment in your own role, and bring what you find back into Session 2.
Use AI on at least three real tasks at work
Real work, not abstract exercises. The kind of thing you actually have to do this fortnight. Aim for thirty minutes a day, experimenting on whatever you are already doing.
Log a thirty second journal entry every time you use AI
What you used it for, how long it would have taken without it, anything you learned. About thirty to sixty seconds each. You can even ask the AI to draft the entry for you based on the work you just did together.
Start your AI Role Plan first pass
Use the GEN Ai AT WORK Role Plan helper. Map your role to its eight key responsibilities and the deliverables under each. Aim for about eighty percent rather than chasing perfection. That is the sweet spot for moving forward.
Bring one example to Session 2
The one that worked best, the one that surprised you most, or the one that did not work and you want help with. We will start Session 2 by hearing two or three of them.
Between sessions. Until Wednesday 20 May.
About 30 minutes a day on your real workAbout 30 minutes a day, on your real work.
Pick something you are actually working on this fortnight. A piece of writing, a meeting summary, a tricky problem you have been sitting with. Try AI on it and see what happens. The aim is not to save time, even though you may. The aim is to build your own sense of what AI does well, where it helps and where it gets in the way, on the kind of work that matters most in your role.
The four specific actions to take across the fortnight are listed in the Actions to take section of the Positioning AI summary above.
Next Session
Knowing What You're Using
Wednesday 20 May 2026 · 2:00 – 4:00 pm AEST · Microsoft Teams