What is AI Fluency?	AI Fluency is the ability to collaborate effectively with AI tools, not just knowing which buttons to click but developing the judgment to use AI well across different situations.
What are the four core competencies of the 4D Framework for AI Fluency?	The four core competencies are Delegation (deciding what work should be done by humans and AI), Description (effectively communicating with AI systems), Discernment (thoughtfully evaluating AI outputs, processes, behaviors, and interactions), and Diligence (using AI responsibly and ethically).
What is the purpose of running evals?	Evals help you understand where Claude adds the most value in your workflow, identify tasks where you'll need to provide more context or examples, and build confidence in Claude's outputs for recurring tasks.
How can you evaluate how well Claude performs on specific tasks that matter to you?	Gather examples of a task you do regularly, create test prompts based on those examples, run your prompts with Claude, compare the outputs to your examples, ask yourself if Claude captures the key information and if the tone and style are appropriate, and refine your approach based on what you learn.
What should you do if AI's response is too generic?	Add details about your specific situation, audience, role, or constraints. Instead of a general prompt like "Write an email about the project delay," try something more specific like "Write an email to our enterprise client explaining that the software integration will be delayed by two weeks."
What is the specific analytical task that Rio wants to delegate to AI?	Cohort analysis
Why does Rio need enrollment dates in the data for cohort analysis?	Without enrollment dates, AI will try to infer them, which he doesn't want.
What should Rio do when using the validated approach with new data?	He'll check whether numbers make sense based on what he knows about his programs, take accountability for the final report, and be transparent about AI's role if asked.
How does the framework for delegating to AI work?	Identify a specific analytical task, be precise about what you need, find past data where you already completed that analysis, work with AI to reproduce your past analysis, systematically evaluate the results, identify gaps, refine your delegation, and then test again.
What can AI help with besides coding tasks?	Brainstorming and implementing solutions you might not have thought of on your own.
Why is validation important when working with AI?	Validation builds confidence in using AI for data analysis but doesn't eliminate responsibility. You're still accountable for checking that the results make sense and being transparent about AI's role in your analysis process.
Can the testing and validation process be used for any type of data analysis or analytical task?	Yes, it works for any type of data analysis, such as donor analysis, budget forecasting, survey synthesis, outcome tracking.
What will the next lesson focus on?	The next lesson will focus on workflow augmentation and how to apply these principles when AI handles routine tasks on your behalf.