Almost everyone has now opened an AI assistant, typed a question, received a fluent paragraph and quietly wondered whether any of it was true. That gap between a confident answer and a usable one is what this day is about. No mathematics, no coding, no predictions about the future of work: six hours on what these tools do, how to get something out of them you would be willing to put your name to, and how to tell when they are wrong.
You start with a plain explanation of the machinery. A large language model produces text by predicting what should come next, learned from an enormous amount of writing. That single fact explains most of the behaviour that confuses people: why it is fluent about subjects it knows nothing about, why it invents a citation that looks perfectly formatted, why it agrees with you too readily, and why it cannot tell you what happened this morning unless the tool you are using has search or a document attached. You see each of these fail in front of you rather than being told about them.
The rest of the day is practice on your own work. You bring three tasks from a normal week, and by the end you have made each of them work: a first draft written from a rough brief, a long document or thread summarised into something a colleague can act on, a set of survey comments or messy notes sorted into themes, an email rewritten for a different reader, and a meeting turned into decisions and actions. You learn to attach your own source material so the assistant works from your facts instead of its memory, and to ask for the output in the shape you need, whether that is a table, a checklist or a five-line reply.
The last part of the day is the part that keeps people out of trouble. You practise a verification habit for anything that leaves your desk: which claims to check, how to make the assistant show where an answer came from, and when to stop and do the task yourself. You work through what should not be typed into a public assistant at all, including customer records, identifiable personal data, unpublished financial figures and anything covered by a confidentiality agreement, and you draw up a short personal rule for your own role. Sri Lankan learners cover this alongside Sri Lanka's Personal Data Protection Act, and teams handling EU or UK personal data alongside GDPR. This is orientation, not legal advice; your own obligations should be confirmed with your legal or compliance contact.
You leave with a personal AI playbook: your own tasks written up as prompts that worked, with the checking step attached to each. The class runs live online in your own timezone or in the classroom in Colombo, with pricing in LKR for learners in Sri Lanka and USD internationally, and the recording available to you for twelve months.
