A.1.1: Case Study Topic Area #1: Mechanics & Application

The Revealing Mechanism

Note 1: This outline provides guidance specific to Case Study #2, complementing the general case study rubric [link].
Note 2: See [link] for some illustrative examples of mechanisms.

The Brief

Understand the high-level algorithmic structure of an AI-powered method, function, or system.

Investigate how it operates and analyse how its design shapes behaviour.

You are not analysing surface outputs.
You are analysing the underlying process.

Your analysis should address the following questions.

1. What problem or function is the system designed to address?

• Problem / Function

    – What is the main objective of the task?
    – Why is this problem important?
    – Why does this mechanism exist within the system?

2. What is the mechanism?

• Mechanism Explanation

    – Describe clearly how it operates.
    – Explain how the overall task breaks down into algorithmic steps or modules.

• Your explanation should show

    – The overall workflow of the system
    – The specific mechanism you are analysing
    – Inputs → transformations → outputs

Avoid unnecessary jargon.
Focus on structural clarity.

3. What does this mechanism enable — and what does it limit?

• Capabilities

    – What strengths or behaviours does the mechanism enable?

• Limitations

    – What trade-offs, distortions, or failure patterns arise from its design?

4. If the mechanism changed, what would change?

• Causal Reasoning

    – Explain how structural design choices create the system’s strengths, limits, and failure patterns.

• Comparison

    – Where helpful, compare your mechanism to an alternative design or configuration.

Your analysis should reveal one key insight about how the mechanism shapes system behaviour.

What a Strong Submission Demonstrates

A strong submission will:

Your work should demonstrate algorithmic thinking and structural clarity.

Choosing Your Mechanism

Your mechanism should be:

Do not choose something broad like “AI” or “machine learning.”
Focus on one clearly defined mechanism within a system.

Explain the mechanism conceptually.
You may reference specific techniques where necessary to clarify how it works.

Do not turn this into a technical manual.

Clarity over breadth.
Mechanism over surface.

See [link] for some illustrative examples of mechanisms.

Deliverables

Your submission is built around a central artefact, which forms the backbone of all three submission components.

The Artefact — Mechanism & Evidence

The artefact is the core analytical and visual material developed through your case study.

Think of it as the body of evidence from which you construct your abstract, slides, and video. It should make the selected mechanism legible by connecting the system context, mechanism, evidence, and resulting behaviour.

Your artefact should include:

• System Context

    – What problem or function is the system designed to address?
    – Where does the selected mechanism sit within the larger system?

• Mechanism Representation

    – Show the overall workflow of the system.
    – Isolate the specific mechanism or module being studied.
    – Trace inputs → transformations → outputs.
    – Where useful, move between system-level and module-level views.

• Evidence & Behaviour

    – Use examples, comparisons, experiments, diagrams, outputs, or other evidence to show what the mechanism enables or limits.
    – Connect structural design choices to observable capabilities, trade-offs, or failure patterns.

Each piece of evidence should be carefully selected, organised, and captioned to explain what it shows and why it matters.

The artefact is not a separate submission format. The same underlying evidence should be used consistently across your abstract, slides, and video.

You may include supporting material such as:

• Diagrams
• Screenshots
• Input / output comparisons
• Prompt or parameter comparisons
• Experimental results
• Tables or quantitative observations
• Code snippets, where relevant

i. Abstract

Provide a concise written abstraction of no more than 300 words.

The abstract should distil:

• The system and mechanism studied
• The question or problem investigated
• The approach taken
• The key finding or behavioural insight
• The principal conclusion about how the mechanism shapes what the system can and cannot do

The abstract should synthesise the strongest argument emerging from the artefact rather than introduce additional material.

ii. Slides

Develop a sequence of portrait-format slides that organises and connects the artefact into a clear visual narrative.

The slides should progressively communicate:

What is the system? → What is the mechanism? → How does it work? → What does it enable or limit? → What is your key insight?

Use diagrams, evidence, comparisons, and concise captions wherever possible.

The slides should not simply collect outputs.
Together, they should construct the visual argument of the case study.

iii. Video

Develop a short-form video presentation of no more than 90 seconds using the slides and artefact as its visual foundation.

Use narration, sequencing, animation, and transitions where useful to connect the evidence and communicate:

• The system or application studied
• The mechanism investigated
• How the mechanism works
• What the mechanism enables or limits
• Your key insight about how its design shapes system behaviour

The video should connect the evidence into a coherent story rather than simply describe each slide in sequence.

iv. Poster

Finally, remember to bring your artefacts together into a complete A2-format poster.

The poster should act as a self-contained summary of the case study, distilling the strongest material from your slides into a single visual argument.

Grading Criteria

In a nutshell

We are not grading:

We are grading:

Criterion What It Means
Understanding Do you correctly explain how the mechanism works?
Causal Reasoning / Comparison Do you link structural design choices to observable system behaviour, including how behaviour might change under alternative designs?
Limitations Do you identify meaningful trade-offs, constraints, or failure modes introduced by the mechanism?
Clarity Can you explain complex ideas in a precise and accessible way?
Visual Explanation Do your diagrams or visual representations clearly communicate the mechanism and information flow?

Rubric

Criterion Excellent Adequate Insufficient
Mechanism Accuracy Structurally correct and clearly articulated explanation of how the mechanism operates Mostly correct explanation with minor gaps or simplifications Superficial, vague, or incorrect description of the mechanism
Causal Analysis Clear linkage between structural design and system behaviour, with thoughtful comparison or reasoning Some connection between structure and behaviour, but underdeveloped Describes outputs or observations without structural reasoning
Failure Awareness Identifies meaningful trade-offs, limitations, or structural constraints Mentions limitations but without depth or explanation No meaningful discussion of limitations
Clarity of Communication Complex mechanism explained clearly and accessibly to non-specialists Explanation understandable but dense or uneven Obscure, overly technical, or difficult to follow
Visual Explanation Diagrams clearly reveal the system workflow, mechanism, and information flow Visuals present but only partially clarify the mechanism Visuals absent, confusing, or decorative rather than explanatory

The Standard

The goal is not to describe AI.
The goal is to understand how it works.

Do not remain at the surface of outputs.
Demonstrate that you can think in mechanisms, modules, and pipelines.

Don’t just use AI.
Explain it.