AI is on most work laptops now, but few people were shown what it is. This free course is the front door, and it teaches through one story. Halvorsen Home Comfort, a fictional 212-person heating and plumbing company in Columbus, Ohio, has three weeks before a cold snap and a CEO who wants an AI plan announced at the next all-hands. You follow the people doing the work: a COO weighing an "AI receptionist" that claims 95% of calls booked, a customer care manager running an eight-day drafting pilot against an inbox that hit 1,140 unanswered messages last winter, and checking whether it actually prevents a repeat, a marketer whose AI-written post promised a federal tax credit that had already ended, and an office manager who learns a rep pasted a customer's garage code into a personal chatbot.
Every lesson works a real problem: the actual request, the actual reply, what an expert would change, a failure and its fix, and a short exercise with a model answer. You learn the difference between rules, prediction models and generative AI; why next-token prediction makes a chat assistant fluent but not always right; a three-part fit test and how to report a result honestly (baseline, denominator, what it leaves out); how to tell a hallucination from an out-of-date answer or a biased pattern; a four-part way to ask; and a one-page rule for what goes into which account, whether that is Microsoft Copilot Chat, Gemini in Google Workspace or a personal app. Public cases, from an FTC action to court sanctions, show the same mistakes outside the story. Three five-minute Decision Labs let you make the calls yourself and see the consequences, and a final lesson turns the method into a one-page plan for a task from your own job. Plan on about an hour for the lessons and labs, plus a 15-minute, 10-question scenario exam that earns a free Wirewalk Institute certificate of completion and assessment.