Getting value from AI at work comes down to three everyday habits: ask clearly, check carefully, and learn from what has already gone right and wrong for other people. This beginner course builds all three. In session one, trainees learn the four elements of a clear request (context, instruction, structure, and audience) and four simple techniques for improving a weak answer. In session two, they learn the five kinds of AI error, a five-step evaluation, and practical ways to verify an answer, including how to tell whether a source the AI cites actually exists. In session three, they analyze workplace AI cases, from a hotel chatbot to resume screening and invoice automation, plus well-known public incidents, and turn the lessons into a safe five-step workflow for one task in their own job. Labs use fictional Massachusetts employers such as Cape Harbor Inn, Merrimack Freight & Logistics, and Bayside Precision Parts and run in the employer's approved tool or on paper. Trainees leave with a prompt makeover log, an evaluation record, a case analysis, and a personal safe AI workflow card. Built on Laurence Lars Svekis's book AI Literacy for All Employees (chapters 6, 7, 13, and 15). No technical background is needed.
What you will be able to do
By the end of the course, trainees will be able to:
- Write a workplace prompt that includes context, a clear instruction, a structure requirement, and the audience, and improve the answer with at least two follow-up prompts.
- Name the five types of AI output error (factual, incomplete, context misread, logical inconsistency, bias or imbalance) and find at least four of five seeded errors in an AI draft.
- Apply a five-step evaluation (accuracy, completeness, relevance, bias, judgment) and at least three verification techniques, including checking that every cited source exists and says what is claimed.
- Analyze a workplace AI case for benefits, challenges, lessons, and the safeguards that would have changed the outcome.
- Write a five-step safe AI workflow (define, instruct, review, judge, document) for one recurring task in their role, with a short AI-use note.
Who this course is for
Office and administrative support; Customer service representatives; Production and warehouse workers and leads; Sales and retail associates; Front desk, scheduling, and patient access staff; Bookkeeping and accounting clerks
Before you start
None beyond basic computer use (email, web browser, word processor). No prior AI use is needed. Open to every W-2 employee, from the shop floor to the front office.
Hands-on activities
- Prompt makeover: pairs turn three weak prompts from fictional Cape Harbor Inn into four-element prompts and compare the answers.
- Follow-up ladder: trainees improve one answer with step-by-step, iterative, role, and constraint follow-ups.
- Error hunt: teams find five seeded error types and one invented citation in an AI-written delivery report for fictional Merrimack Freight & Logistics.
- Case clinic: small groups analyze one of five workplace cases and present the safeguard that would have changed the outcome.
- Safe workflow build: each trainee writes and peer-tests a five-step workflow card for a real task type.
How your learning is measured
- Attendance at all three sessions (start and end check).
- Lab worksheets W1 to W3 scored complete / incomplete against written criteria.
- Pre/post knowledge check from the 48-item quiz bank; gain reported per trainee; 80% on the final draw to meet.
- Final W4 safe AI workflow card scored Meets on every rubric criterion (one resubmission allowed).
- Pre/post confidence survey (reported, not graded).
