The Ghost in the Machine: AI Resource pack

The Ghost in the Machine: A Hands-On Guide to AI, Logic, and Ethics
Are your students curious about how ChatGPT actually “thinks”? Move beyond the hype with this comprehensive mini-unit designed for A/AS or GCSE Computer Science. This resource pack takes a deep dive into the mechanics of Large Language Models (LLMs) through live Python coding, data-driven experiments, and exam-style evaluation.

What’s Included:

Slide Deck: “Demystifying the Next Token”: An engaging PowerPoint that breaks down complex concepts like Tokenization, Encoding, and Statistical Prediction into student-friendly analogies.

Simple_Python_SLM: A simple, elegant script that demonstrates a “Markov Chain” in action. Perfect for teaching basic list logic and dictionary mapping.

Python_SLM_AI: An upgraded script using File Handling (.txt persistence) and Text Pre-processing. Students learn how to build a model that “remembers” what it has been taught over time.

The “Bias Lab” Dataset: A specially curated, “synthetic” news report designed to bake specific gender biases into the Python model. This provides a powerful “Aha!” moment when students see their own code generate discriminatory outputs based on the training data.

Exam board-Style Assessment Pack: * Four original 8-mark extended response questions covering Ethical, Legal, Cultural, and Environmental impacts.

PEEL Writing Frames and Success Checklists to support literacy.

Model Answers: Side-by-side “Good” vs. “Poor” examples with marking commentary to help students understand the OCR Level of Response criteria.

Key Learning Objectives:

Explain how LLMs use tokens and statistical probability to generate text.

Implement a functional Markov Chain using Python dictionaries and File I/O.

Critically evaluate the impact of algorithmic bias on society (Mortgages, IP law, and the Environment).

There is even a keyword Bingo card to help keep the students engaged.
Perfect for a 2–3 lesson “Deep Dive” or a high-impact enrichment week!