Prompt Engineering with ChatGPT: Get Reliable, Repeatable Results

ChatGPT has become an essential tool for knowledge work, but getting reliable, high-quality results takes more than typing in a question and hoping for the best. In this beginner-friendly course, learn how ChatGPT works and how to prompt it effectively using a repeatable playbook. Practice using seven core techniques: adding context, being specific, using roles and personas, breaking complex requests into smaller steps, applying custom instructions, working with images and other inputs, and using Socratic questioning to improve accuracy. Explore advanced features such as choosing the right mode and model, organizing work with projects, and building custom GPTs. This course also covers security essentials, including what not to share, how to control your privacy and data settings, and how to spot and reduce hallucinations. Along the way, hands-on challenges give you a chance to start applying your newly acquired skills right away.

Learning objectives
Identify the key elements of an effective prompt (context, specificity, constraints, and format) and apply them to produce more accurate, useful outputs.
Demonstrate advanced prompting techniques (personas/roles, task decomposition, Socratic questioning, and custom instructions) to improve the clarity, tone, and completeness of your results.
Select the appropriate ChatGPT feature or workflow (mode/model choice, projects, custom GPTs, and metaprompts) for a given work scenario.
Lower ChatGPT risks by identifying sensitive information, configuring privacy/data controls, and evaluating outputs to detect and mitigate hallucinations.
Analyze when and why ChatGPT outputs may be unreliable, and identify practical techniques, like verification prompts and source checks, to improve accuracy and trustworthiness.