<aside>

For Claude Certified Associate - Foundations

For Claude Certified Architect - Foundations

</aside>

The 4D Framework

Four interconnected competencies for effective, ethical AI collaboration (developed by Anthropic with Prof. Rick Dakan and Prof. Joseph Feller):

Competency	What it means
Delegation	Deciding what work goes to AI vs. stays human. Covers Problem Awareness, Platform Awareness, Task Delegation.
Description	Communicating clearly with AI: defining outputs, guiding process, specifying tone/behavior. This is "prompt engineering" reframed as a general professional skill.
Discernment	Critically evaluating AI outputs, process, and behavior; assessing accuracy, quality, appropriateness.
Diligence	Taking responsibility for AI-assisted work: transparency, accountability, ethical use (Creation, Transparency, Deployment diligence).

Weak Delegation breaks the whole chain, no amount of good prompting saves a poor delegation decision.
Maps to 3 modes of working with AI: Automation (AI executes a defined task), Augmentation (iterative human-AI collaboration), Agency (AI operates more independently toward a goal).
The Description-Discernment loop: you describe, evaluate the output, refine your description, repeat.

Effective Prompting Techniques (6 foundational)

Give context: be specific about scope, background, and goal.
Show examples: demonstrate the format/style you want (few-shot).
Specify constraints: define length, format, and other output requirements.
Break complex tasks into steps: guide multi-step reasoning explicitly.
Ask the AI to think first: give it space to reason before answering.
Define the AI's role or tone: sets perspective, depth, and communication style.

Secret weapon: ask Claude itself to help you improve your prompt.