Workshop – Getting Things Done With AI January 2026

Getting Things Done with AI

Date and location:

When: Wednesday, January 7th
Time: 10:00 am – 11:15 am
Where: Law Library 300 & Remote Option via Zoom
Who: Everyone

The first in a new ongoing series of lab-style sessions is designed to help you put AI to work on your tasks — teaching, admin, planning, writing, brainstorming, process cleanup, and more. In this session we’ll workshop prompts and approaches together, then you’ll spend most of the time building and iterating on real workflows.

Future sessions in this series will support things like:

  • Identifying and prototyping potential high-value use cases
  • Refining and systematizing repeatable routines with AI
  • FAQ and troubleshooting sessions

Feel free to bring a new idea each month or refine work from prior sessions.

Workshop Resources

Detailed Summary of Topics Covered

Workshop Overview: “Getting Stuff Done with AI”

This hands-on workshop introduced faculty and staff to effective collaboration with AI tools through practical experimentation and real-world task application. This summary was created by the “transcript-processing” skill created by Kyle Fidalgo and used with Claude Code. Content was reviewed for accuracy by a human.

Core Conceptual Frameworks

The Five Design Basics for Working with Generative AI

  1. Sentient Design Material: AI is a probabilistic, highly adaptable tool that’s context-aware, multimodal, and flexible for knowledge work
  2. Iteration as Real Work: Working with AI requires continuous refinement – it’s simple (just typing) but not always easy (requires thoughtful prompting)
  3. Meta-Prompting: When stuck, ask AI how to work better with it – use it as both vehicle and guide
  4. Context is Everything: AI needs your goals, audience, success criteria, and relevant background information to be effective
  5. Natural Language Interface: If you can write clearly, you can collaborate with AI – no technical expertise required

The Four Ds of AI Fluency (Anthropic Framework)

  1. Delegation – Knowing what to hand off to AI vs. keep for yourself vs. co-create
  2. Description – Clear prompting and context engineering (good communication = good prompting)
  3. Discernment – Critical evaluation of outputs for accuracy, quality, relevance, and ethical considerations
  4. Diligence – Taking ownership and responsibility for AI-generated output, including proper attribution

Effective Prompting Principles

  • The Golden Rule: Could a grad assistant or new colleague complete this task successfully with just your prompt and no additional context?
  • Provide Appropriate Context: Background information, audience, use case, success criteria, formatting expectations
  • Think Like Onboarding a Brilliant Employee with Amnesia: Each new chat requires full context – the AI has no memory of previous sessions
  • Avoid Over/Under-Contexting: Too little = confused AI; too much = AI focuses on wrong things

Prototyping Mindset

  • Permission to fail boldly – learn from both successes and failures
  • Think out loud into the text input
  • Don’t overthink – just try something
  • Good enough is often better than perfect when learning
  • Use AI on real tasks you care about to build muscle memory

Advanced Concepts Introduced

  • The Harness Concept: What tools and capabilities does your AI model have access to? (e.g., basic chat interface vs. agentic systems like Claude Code that can autonomously research, write code, and execute tasks)
  • Description-Discernment Loop: The iterative cycle of describing tasks, evaluating outputs, and guiding the AI back on track
  • Edit/Re-roll Prompts: When AI makes mistakes, edit the problematic prompt to fork a new conversation path rather than continuing with errors in context
  • Break Down Complex Tasks: Divide larger projects into discrete, manageable steps

Practical Applications Demonstrated

  • Task planning and organization (hub-and-spoke project management model)
  • Building custom tools (workshop timer app with Claude Code)
  • Research and fact-finding missions
  • Template and framework creation
  • Decision trees and checklists
  • Process documentation for future reference
  • Case analysis and legal research challenges

Key Takeaways

  • Skilled AI use is about being an effective communicator and collaborative partner, not understanding LLM technical details
  • Always be experimenting (ABE) – use AI on real tasks you care about
  • Share learnings and document what works for you
  • Stay curious as the technology evolves rapidly

Quick Tips for Putting This Into Practice

Getting Started

  1. Pick one real task – Choose something personal or work-related you actually need to do (not a hypothetical)
  2. Start with the interview approach – Ask AI to interview you about your task to help build context together
  3. Use the Golden Rule test – Before hitting send, ask: “Could someone unfamiliar with my work complete this task with just my prompt?”

When You’re Stuck

  1. Meta-prompt – Ask the AI: “What additional information do you need from me to help with this task?” or “What am I doing wrong in how I’m describing this?”
  2. Break it down – If the AI is struggling, simplify the scope or divide the task into smaller steps
  3. Edit, don’t continue – If the AI makes a mistake, use “Edit prompt” to fork a new path rather than trying to correct within the same thread

Building Better Prompts

  1. Provide the 4 W’s: Who is the audience? What does success look like? Why are you doing this? What workflow is this part of?
  2. Be explicit about format – Tell AI exactly what structure you want (bullet points, table, narrative, etc.)
  3. Give examples when possible – Show AI what “good” looks like for your specific context

Avoiding Common Pitfalls

  1. Don’t trust without verifying – AI can “hallucinate” plausible-sounding but incorrect information, especially with specialized knowledge
  2. Put source material in the conversation – Don’t ask AI to recall facts from memory; give it the documents to work from
  3. Check what tools your AI has access to – Basic chat interfaces have limitations; agentic systems can do autonomous research and execution

Building the Habit

  1. Always Be Experimenting (ABE) – Regular practice builds intuition for when AI is/isn’t helpful
  2. Document your wins – Save effective prompts and patterns (use Google Gems, Claude Projects, or ChatGPT custom GPTs)
  3. Share and learn together – AI literacy grows faster through community teaching and experimentation

Tool Recommendations

  • Google Gemini (preferred at BC): Use Gemini 3 Pro model for serious work
  • Advanced workflows: Explore agentic tools like Claude Code for multi-step autonomous tasks

Learning resources: AI Foundations (Kyle’s site) or Anthropic’s AI Fluency course