ChatGPT
Use large language models well \u2014 prompting, workflows, limits.
- Self-paced
- 8–10 weeks
- 25+ hrs
- Credential included
Curriculum
5 subjects · 12 chapters · 48 topics
- 01
Understanding the Tool
1.1 How ChatGPT works
- Language models in plain terms
- Tokens, context windows and limits
- Why it makes things up
- Where it is strong and where it is not
1.2 The interface
- Chats, projects and history
- Custom instructions
- Files, images and voice
- Choosing between models
- 02
Prompting Well
2.1 Prompt fundamentals
- Being specific about task, format and audience
- Giving examples
- Assigning a role and constraints
- Iterating rather than restarting
2.2 Advanced prompting
- Step-by-step reasoning
- Breaking a big task into chained prompts
- Templates you can reuse
- Getting structured output
2.3 Common failures
- Vague prompts and vague answers
- Hallucinated facts and citations
- Losing the thread in long chats
- How to verify an answer
- 03
ChatGPT for Work
3.1 Writing
- Drafting emails, reports and proposals
- Editing and tightening your own writing
- Matching a tone of voice
- Summarising long documents
3.2 Analysis
- Working with spreadsheets and CSVs
- Asking questions of a document
- Making comparison tables
- Building simple charts
3.3 Research and study
- Explaining a difficult topic
- Building study plans and flashcards
- Extracting key points from a paper
- Checking what you have understood
- 04
Building with ChatGPT
4.1 Custom GPTs
- Defining purpose and instructions
- Adding knowledge files
- Configuring actions
- Sharing inside a team
4.2 The API
- API keys and first request
- Roles, messages and parameters
- Cost and token budgeting
- Simple automations
- 05
Using It Responsibly
5.1 Accuracy and trust
- Fact-checking a response
- Knowing when not to use it
- Citing sources honestly
- Academic and workplace policies
5.2 Privacy and safety
- What not to paste into a chat
- Company data and confidentiality
- Data controls and retention
- A capstone workflow of your own
