Critical Thinking for AI

This introductory-level program provides a plain-language explanation of how Large Language Models (LLMs) and the AI tools built on them operate. Through scenarios that reveal the true capabilities and limitations of AI tools, this program builds critical thinking skills that support their appropriate application in the workplace. This program is based on the expertise of Aaron Barth, PhD. Combining an academic background in philosophy with science, smart design, and technology, Dr. Barth works to change behaviour and transform organizations. He has delivered critical thinking workshops across North America for over fifteen years.

TOPIC: Collaboration

LEVEL: Introductory

Core Skills & Behaviours

Assess risk and judgment involved in a task: Evaluate tasks based on the impact of failure to determine when AI assistance is appropriate and when it poses an operational liability.
Evaluative AI outputs: Identify sycophantic language and superlatives in AI outputs, rigorously verifying the underlying logic, accuracy, and applicability.
Mitigate AI bias: Recognize and correct for the exclusion of contested or non-dominant views, statistical outliers, and niche domain expertise in LLM-generated responses.
Protect skills and cognitive health: Prevent cognitive offloading to acquire and preserve core professional skills. Keep some high-value tasks AI-free to manage mental fatigue from high-volume AI outputs.
Consider impacts: Critically examine new AI implementations through an equity lens, identifying groups who may bear disproportionate impacts from automation to determine if it is beneficial and appropriate.

	Snippet Title	Topic Covered
1	The Safe Bet	Encourages learners to break down tasks and assess the level of risk and judgement required before delegating to AI chatbots.
2	The Grand Opening	Explains how AI tools based on Large Language Models (LLMs) mimic arguments but do not reason or “think” logically. Suggests techniques for reviewing logic.
3	A Strange Trend	Demonstrates how LLMs prioritize Euro-American perspectives and erase others. Recommends the inclusion of human specialists when non-dominant information is important.
4	Get This Right	Explores the intellectual and professional consequences of cognitive offloading to AI tools. Suggests strategies to use AI as a thinking partner.
5	The New Consultant	Discusses the limits of AI’s “intelligence” and sets appropriate expectations about what it can and cannot do.
6	The “Masterclass” Proposal	Explores the problems of sycophancy, weasel words, and empty rhetoric that impact AI’s ability to act as a critical reviewer. Suggests prompting techniques to minimize these tendencies.
7	Same Job, New Tasks	Explores the concept of cognitive overload from AI use and suggests modifying workflows to preserve periods of deep work.
8	Cutting Costs	Explores AI’s inability to understand context and underscores the need for professional judgement.
9	Thirty Slides of Nothing	Introduces the concept of AI-generated “workslop” and provides steps for addressing the problem respectfully with team members.
10	Build a Bot	Examines the gender-equity risks of automation and provides a framework for evaluating the human impact before implementing new AI tasks.

Critical Thinking for AI

TOPIC: Collaboration

Try Learning Snippets FREE