Leon Furze

This is the fourth post in a series exploring the AI Assessment Scale (AIAS) in more detail. In the previous posts, I looked at Level 1-3. This time we head towards a balanced use of AI and explore how students might deliberately incorporate Generative AI into their work at all levels.

Just a reminder if you’ve missed any of this so far: the AIAS can be used before, during and after assessment and is a tool to provide clarity for both students and educators. 

It’s not a way to “catch” or detect AI use: there are no foolproof methods to tell whether a student has used AI or not. Rather, the AIAS allows for a conversation with students about the appropriate use of

Level 4: AI Completion, Human Evaluation

Tasks in Level 4 make the most of the emerging capabilities of Generative AI, but still require human judgement and oversight. For most tasks, we’re not yet at the stage where AI can be used unsupervised. The tendency for Large Language Models to hallucinate, for example, or the obvious bias in image generation, mean that students need to exercise caution when trying to generate entire responses. 

There are also opportunities for students to engage critically with GenAI systems, ensuring that they are using the technology consciously and in ways which support their learning. 

There are a few reasons why you might want to encourage a 50/50 use of GenAI, such as:

• AI generation of resources and data: The use of GenAI to create specific content or datasets, for use in a longer project. 
• Comparative analysis: Comparing AI-generated content with human-created work, for example to identify missed ideas or opportunities in either work.
• Critical Evaluation: Critiquing AI content to assess biases, inaccuracies, and other limitations. 
• 50/50 use: Incorporating AI content into larger projects, for example generating a lot of content (50% or more) using image generation in a digital arts assessment 

And a couple of suggestions for running activities at this level:

• The human author is still responsible for the final submission
• A complete submission generated by AI might be submitted alongside the human submission

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 4 Assessments

Here are some ideas for tasks at Level 4 which require AI generation and human oversight:

1. English: Generate draft arguments for the purpose of critiquing and interrogating the machine output.
2. Physics: Use GenAI as a ‘data analyst’ to help analyse and visualise data, before checking manually for inconsistencies or errors.
3. Politics: Side-by-side comparisons of AI-generated political campaigns modelled on real political campaigns.
4. Engineering: AI-designed prototypes, including image generation, which students then test and refine.
5. Environmental Science: AI used to model climate change effects, with students proposing mitigation strategies.
6. Psychology: AI to simulate patient scenarios for diagnostic and treatment planning exercises, while the human student assesses the overall quality of the scenarios.
7. Drama: AI to draft scripts for individual characters in plays written by students, which students then develop into full productions.
8. Hospitality: Use GenAI to suggest recipe variations, which students execute and critique.
9. Fashion Design: Use AI such as image generation to propose clothing patterns, with students creating and adjusting the final designs.
10. Art and Design: Students use AI to generate artwork or design concepts, followed by a comprehensive critique of the generated piece, assessing its aesthetic, cultural, and technical aspects and any evidence of bias.

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 Monday February 12th.

Lesson idea: Critical AI Folio

This activity introduces students in the Visual Arts to image generation, and offers a way to use the technology both critically and creatively. Students generate a folio of digital artwork using the technology, then discuss its strengths, limitations, and ethical considerations. They also review some of the more problematic aspects of GenAI and then decide for themselves how much of the the technology to incorporate into their final piece, which can be up to 50% generated by AI.

Activity Outline:

1. Introduction to GenAI in Art:
   • Brief students on GenAI and its current place in visual arts.
   • Discuss issues surrounding the inability to copyright GenAI artworks according to current laws worldwide.
   • Discuss at the ongoing copyright and IP issues with datasets, such as articles about the UK artists used in training data.
2. Exploration and Research:
   • Assign further reading on the complexities and ongoing legal discussions surrounding GenAI and copyright.
   • Guide students to relevant articles that outline the current legal stance and debates on AI-generated art and copyright issues.
3. Folio Creation:
   • Students use GenAI to generate initial pieces of digital artwork. Use a variety of platforms if possible, such as Copilot (formerly Bing Image Creator) or Adobe Firefly.
   • Discuss the strengths and limitations of the technology, and whether the ease of use balances the issues researched earlier.
   • They then decide to add to or modify this AI-generated folio with their own digital artwork, ensuring that their final piece is at least 50% their own work. They may use “traditional” digital tools such as Adobe Photoshop, with or without the AI features such as generative expand/fill.
4. Critical Analysis and Discussion:
   • Host a critique session where students present their folio and discuss the role of GenAI in their creative process.
   • Encourage debate on the ethical implications of using GenAI in art, referencing specific examples from the provided readings and their own experiences with the technology.
5. Final Piece Submission:
   • Students submit their final artwork, along with a written reflection on their use of GenAI, the creative choices they made, and their stance on the copyright issues discussed.

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: