Leon Furze

This is the fifth and final post in a series exploring the AI Assessment Scale in more detail. The previous posts can all be found here:

• Level 1: No AI
• Level 2: Ideas and Structure
• Level 3: AI Editing
• Level 4: AI Completion, AI Evaluation

Level 5: Full AI

This is the final level of the Scale and the one which permits the most use of AI – essentially, students can use any GenAI tool while completing the assessment. This level, and the previous with a 50/50 split of human/AI, is probably the most similar to how students will use AI “out in the wild”. After all, it’s not likely that many employers will police AI use to the extent that happens in education.

There are many reasons why you might want to permit full AI use in any task, for example:

• Use of AI is encouraged, or is the active focus of the task (i.e., students are taught how to use GenAI tools for the purpose of assessment)
• The assessable skills are outside of the context of AI (e.g., discussion, collaboration, practical work) and therefore it does not matter if AI is used
• AI use is encouraged in ways which are novel and creative
• AI use is encouraged in ways which are “administrative” (e.g., recording the minutes of group discussions) and applicable to use outside of the classroom/education context

Practical AI Strategies includes an entire section on GenAI policy and assessment. It is available from Amba Press

Enjoying the article and want to stay up to date with the rest of the series? Join the mailing list:

Ideas for Level 5 assessments

Throughout this series of posts I have tried to present a diverse range of assessment task ideas across different subject areas. At Level 5, all of these assessment tasks permit unrestricted use of GenAI tools such as text, image, code, and audio generation as well as related tools like automatic speech recognition (ASR).

1. Computer Science: Students use GenAI tools like Github Copilot to develop and debug software projects, leveraging AI for code generation and testing.
2. Business Studies: Creation of AI-generated market analysis reports, including predictive modelling of market trends using GenAI tools for data analysis and visualisation.
3. Mathematics: Using code interpreter style functions in GenAI to solve complex mathematical problems and proofs, with students then explaining the AI’s approach and verifying its accuracy.
4. Literature: Writing entire texts or short stories using GenAI, with students focusing on editing, refining, and adding depth to characters and plot lines generated by AI for the purpose of exploring the quality of GenAI output, encoded biases, etc.
5. Biology: Using GenAI to create entire datasets for analysis as practice or mock experiments while learning processes, or using GenAI to write scripts (e.g., in Python) to create tools for data analysis, fake datasets, etc.
6. Music: Composition of full musical pieces using GenAI, with students arranging, producing, and critiquing the AI-generated music for technical and emotional depth.
7. Film and Media Studies: Students use GenAI to script and storyboard short films or advertisements, focusing on narrative development and visual storytelling.
8. Humanities: Analysis of AI-generated social simulations to explore sociological theories and concepts, with students critiquing the simulations’ assumptions and conclusions.
9. Health and Physical Education: Using GenAI to create training and diet plans, with students evaluating their effectiveness and making adjustments based on health and fitness goals.
10. Geography: AI-generated models of urban development and climate change impacts, with students assessing the models’ viability and proposing sustainable solutions.
11. Graphic Design: Students use AI for the creation of visual designs, logos, and branding materials, focusing on the refinement and application of design principles to AI-generated concepts.

My first webinar for 2024 is Teaching AI Ethics. The session covers key area of ethical concern for Generative AI, including bias, environmental concerns, and copyright. Register below for today’s session!

Level 5 Assessment Task: Advanced GenAI in Biology

This activity is designed for Biology students to make full use of the capabilities of Generative AI (GenAI) in conducting virtual experiments and data analysis. It encourages the use of AI for generating datasets, analysing biological data, and even scripting tools to assist in data manipulation and visualisation. This task not only familiarises students with the technology but also enhances their understanding of biological processes through simulated experiments.

I’m deliberately going for a complex task here to demonstrate what might be possible with GenAI – I don’t expect many Year 11 Biology classes will be incorporating TensorFlow any time soon, but it should serve as an example of how secondary and tertiary educators could think of these tools.

Activity Outline

Training on GenAI Tools and Scripting Languages:

• Provide tutorials on using GenAI platforms suitable for generating datasets and writing scripts (e.g., Python libraries like TensorFlow or PyTorch for machine learning).
• Introduce students to basic Python scripting, focusing on data analysis and visualization libraries (e.g., Pandas, Matplotlib).

GenAI Experiment Design:

• Instruct students to identify a biological research question that can be explored through virtual experiments.
• Guide them to design an experiment where GenAI is used to generate the necessary datasets. This could involve simulating population genetics, ecosystem dynamics, or cellular processes.

Data Generation and Analysis:

• Students use GenAI tools to create datasets for their chosen experiment. This includes specifying parameters for the data generation to reflect realistic biological scenarios.
• Students then write or adapt Python scripts using GenAI to analyse the data, looking for patterns, anomalies, or confirming hypotheses.

Critical Evaluation and Discussion:

• Organise a session where students present their findings and discuss the role of GenAI in their research simulation.
• Encourage a discussion on the reliability of AI-generated data, ethical considerations in using AI for scientific research, and the implications of AI in future biological studies.

Final Project Submission:

• Students submit a comprehensive report detailing their research question, experiment design, GenAI-generated data, analysis scripts, and findings.
• Include a reflective section on the experience of using GenAI in biological research, focusing on the technology’s strengths, limitations, and potential improvements.

Example Prompts for GenAI Platforms

Dataset Generation for Genetic Variation Study:

“Generate a dataset simulating genetic variations across 1000 individuals for a population under selective pressure due to environmental changes.”

Ecosystem Simulation Data:

“Create a dataset representing predator and prey population dynamics over 50 years in a closed ecosystem with specific parameters for birth rate, death rate, and carrying capacity.”

Cellular Process Simulation:

“Generate a dataset simulating the rate of photosynthesis under varying light intensities and carbon dioxide concentrations for a plant species.”

Python Script for Data Analysis:

“Write a Python script to analyse a given dataset for patterns of antibiotic resistance in bacterial populations, including visualisation of trends over time.”

Tool Creation for Genomic Data Interpretation:

“Develop a GenAI-powered Python tool that interprets genomic data to predict phenotype expressions based on genotype information.”

Here are a couple of examples of what these prompts might output in different platforms:

If you have questions or comments, or you’d like to get in touch to discuss GenAI consulting and professional development, use the form below: