How do we use AI on ourselves — to think, reflect and grow — without handing over our judgement?
A critical friend, not a compliant one.
Watch · the keynote trailer
A room that is less afraid and less sold to — practical everyday uses, the skill of prompting for challenge rather than agreement, and a clear line about where a human has to stay in the loop.
Fireside · Conference keynote · Panel · All-hands
About this session
Most AI sessions teach you to be more productive at work — clear your inbox, summarise the meeting, write the report. This one doesn’t. It’s a conversation about using AI on yourself: to think more clearly, to reflect, to prepare for the difficult conversation, to work out what you actually believe before you say it out loud.
That’s a different question, and it matters most to the people who have somewhere to get to and nobody obvious to ask. The teenager who can’t ask their parents. The colleague who doesn’t want to raise it with their manager. The person weighing something up at 2am who just needs to hear their own thinking said back to them.
It runs as a fireside conversation rather than a talk — host-led, unscripted, shaped around a few agreed pillars, with the audience invited in throughout. Joanne is an AI realist: neither evangelist nor doom-monger. The whole point is that the sceptics in the room get a genuine hearing, because a session that only platforms the enthusiasts convinces nobody.
What we explore
The everyday unlock. Start where the fear is lowest. Dictating a rambling account of a car accident and getting a clean insurance claim out of it. A complaint to a rail company that actually paid out. Blood test results explained in plain English. None of this is clever — that’s the point. It’s the on-ramp.
Rubber ducking. The technique comes from software teams who kept a rubber duck on the desk: before you’re allowed to ask the senior developer, you have to explain the problem to the duck. A large share of problems solve themselves in the telling. AI is an interactive duck — it answers back.
Prompt for challenge, not validation. The real risk isn’t hallucination, it’s agreement. Ask it to back you and it will. The skill is telling it to find the holes in your argument — to be honest, fair, not brutal, encouraging but realistic. The friend who tells you that you’ve got bad breath, not the one who lets you walk into the room.
Reflection, with the line drawn clearly. Journaling, and reading your own patterns back weeks later. Thinking through a decision you’re circling. This is where the caveats have to be loudest: it is not a therapist, it is not clinical advice, and there are moments that need a human being. Knowing where that line sits is part of the session, not a disclaimer at the end.
The new digital divide. Twenty years ago we worried about who had a computer. Now a colleague of thirty years hears “agents” and “Copilot” and quietly decides this isn’t for them. AI fluency is becoming a workplace competency — which means it can be taught, and nobody should be written off for not having arrived with it.
Keeping the human in the loop. Where does the hand-off happen? What can you delegate before it stops being yours? The first rung of the career ladder is being automated away — you can’t become a senior anything without having been a junior one. Worth facing squarely rather than cheerfully.
Think, feel and act differently
Think differently — AI doesn’t make you better; it reveals what you bring. The same recipe produces a masterpiece or something merely edible depending on who’s holding it.
Feel differently — less afraid, and less sold to. Curious enough to experiment, sceptical enough to keep asking questions.
Act differently — one thing to try this week, and one boundary you won’t cross.
Who this is for
Employee networks and ERGs, belonging and inclusion conferences, all-hands and lunch-and-learns — anywhere people want to talk about AI as humans rather than as users. Especially strong for communities where “who do I ask about this?” has a complicated answer, and for organisations part-way through an AI rollout where some colleagues are flying and others have gone quiet.
Works best with a named host and a live audience — polls per pillar to temperature-check the room, and questions taken throughout rather than saved to the end.
Go deeper
Bespoke to your audience — but well-trodden. Delivered for LEGO · NHS England · Unilever · John Lewis · lululemon.
“Working with Joanne on the 1st French Tech DIB Pan-European Summit was an immensely pleasurable and enriching experience. Her exceptional talent, innovative ideas, and marvellous sense of humour made her an outstanding partner and advisor throughout the event. Joanne's ability to cause 'good trouble' — the best kind — truly set her apart as a unique and invaluable contributor in the field.”
