Think Better with AI
AI will give you a confident answer to almost anything. This is a day on asking better questions and knowing when to trust what comes back.
About this session
Most people who use AI have had the same experience. You ask a loose question and get back a polished, confident, well-organised answer. It reads well and sounds right. Whether it is right depends almost entirely on what you brought to it: how clearly you framed the problem, what you told it, and what you thought to check.
The good news is that you already know how to do this. Anyone who can solve for x and explain each step is thinking critically. The skill you already have isn't what slips. What slips is noticing whether you are still using it once a fluent answer is on the screen. That noticing is what this day is built around.
You'll work on realistic, ambiguous problems with an AI assistant and learn a five-step routine called CLEAR: Clarify the outcome, Load the context, Examine the assumptions, Assess the output, Refine and reflect. Roughly a third of the day is instruction. The rest is practice, discussion and revision, with two instructors in the room. In the afternoon you apply all five steps to a case you bring from your own work or studies.
It works with whichever AI tool you already use, and it will still work with the one you use in five years.
What you'll leave with
- CLEAR, a five-step method for working any problem with AI, on a one-page reference sheet
- Your own case worked through all five steps, with feedback from the room
- A before-and-after of your work from the day, showing what changed and why
- The habit of separating what you know, what you are assuming, and what you still need to find out
- A way to spot unsupported claims in an AI answer and decide what needs checking before you act on it
Who it's for
- Anyone who uses AI at work or in their studies and wants to trust it for the right reasons
- Students and people early in their careers who want the habit in place before a job depends on it
- Experienced professionals and managers who want a method they can use themselves and pass on to a team
- Complete beginners. No coding and no prior AI experience are assumed
Agenda
- Cold start One ambiguous problem, one AI assistant, a few minutes. We save what you produce and come back to it at the end.
- Asking well Clarify the outcome and load the context: turning a vague request into a usable one.
- Finding the flaw Assumptions, alternative explanations, and deciding what needs to be verified.
- Lunch Provided.
- Noticing how you decide to trust Predicting, testing and revising, and paying attention to what changes your confidence.
- Capstone: your own case Take the case you brought through all five steps, get feedback from other participants, and refine it.
- Close Before-and-after, and one thing you'll do differently next week.
Who's presenting
-
Dr. Dominik Kalisch
Read full bio
-
Jack Byers
Read full bio
Questions people ask
Do I need to know anything about AI?
No. The workshop is built for beginners. If you've never used an AI assistant, we'll get you set up in the first few minutes. If you use one every day, you'll get more out of it by the afternoon.
What should I bring?
A laptop, access to an AI assistant (any one; a free account is fine), and one real case from your work or studies that you're happy to discuss with a small group. Nothing confidential. If you can't bring a laptop, reply to your confirmation email and we'll pair you up.
What counts as a case?
A decision, a problem or a piece of work you actually have to deal with. For example, a report you need to write, a process that isn't working, or a choice you're weighing up. Half a page of notes is plenty. We'll send guidance a week or so before.
Which AI tool do you teach?
None in particular. The method doesn't depend on the tool, which is the point.
What does it cost?
$199 if you register by Saturday, October 10. After that it's $249, or $199 for students who register with a school email address ending in .edu. Please bring a student ID; we check it at sign-in. Lunch and coffee are included. You pay by card when you register, and your receipt and invoice come by email.
What if I can't make it?
Cancel more than fourteen days before and you get a full refund. Between seven and fourteen days before, you get half back. Inside seven days, we can't refund, but you can hand your seat to someone else or move to a future session. To cancel, reply to your confirmation email.
Why only twenty-four seats?
Most of the day is practice, and two instructors can only give proper feedback to so many people.
Will there be photos?
We'd like to take a few during the day. You can opt out when you sign in, and we won't publish a photo where you can be identified without asking you first.
Who is this not for?
If you already treat every AI answer the way you'd treat a confident colleague's first draft, meaning you check the claims and question the assumptions before acting, you probably don't need this day.