Teachers do not have the time to read entire documents to build AI literacy. Dense research papers, policy briefs, and technical explainers create a barrier — the subject matter is important, but the format is inaccessible within the realities of an educator’s working day.

Simulations offer a different path. Rather than asking teachers to read about how AI processes text, simulations let them experience it directly — adjusting inputs, observing outputs, and drawing their own conclusions. Within the education context, this approach converts passive information into active understanding without demanding prior technical knowledge.

Edu AI Lab is a suite of browser-based simulations built on this premise: that the fastest route to AI literacy for educators is not reading, but doing.

Simple interfaces with UI controls that alter AI text output

Each simulation is built around a single, legible control surface. Sliders, toggles, and dropdowns directly alter AI behaviour in real time — switching between word-for-word interpretation, analysis, and review modes — so the relationship between user input and AI output is immediate and visible. There is no abstraction layer between what the teacher adjusts and what the simulation does.

Clear table format, headers, and tooltips to reduce visual clutter

Outputs are structured as tables with clear column headers rather than unbroken prose. Tooltips surface definitions and explanations only when needed, removing the cognitive overhead of heavy explanatory text that sits on screen regardless of whether the user needs it. The goal is an interface that teaches through use, not through reading.

AI Auto Marker

Low-fidelity wireframe exploring the layout and control hierarchy for the auto-marking simulation.

Notes Interpreter

Low-fidelity wireframe for the notes interpretation simulation.

AI Auto Marker

Pros. Demonstrates AI grading behaviour concretely and interactively. Table output makes scoring logic transparent. Real-time feedback loop between teacher input and AI output builds intuition quickly.

Cons. Simplified marking criteria may not reflect the complexity of real classroom rubrics. Risk of over-reliance on AI scores without contextual judgement from the teacher.

Notes Interpreter

Pros. Mode switching (interpretation vs. analysis vs. review) gives educators a direct comparison of how framing affects AI output. Structured table output is easier to scan than a wall of AI-generated prose.

Cons. The quality of output depends heavily on input quality — educators using poorly structured notes may draw inaccurate conclusions about AI capability. Tooltip-based explanations require the user to seek context rather than having it surface proactively.

AI Auto Marker

High-fidelity screens showing the final visual design of the auto-marking simulation: the input and configuration panel, and the structured results output.

Notes Interpreter

High-fidelity screens showing the final visual design of the notes interpretation simulation across its three primary states.