In Lesson 4, you encountered sustainability indicators in the context of bioenergy. Lesson 7 returns to sustainability indicators and applies them directly to your final project topic.
Sustainability indicators are used to make selected conditions, impacts, and changes visible. An indicator can support comparison, evaluation, monitoring, or decision-making. Indicator selection is also an ethical process. The selected indicators determine which benefits, burdens, risks, stakeholders, timescales, and system functions receive attention. Other concerns may remain difficult to measure or may disappear from the analysis.
During this one-week lesson, you will use generative AI as a structured analytical assistant. You will provide the same project context and ethical criteria to two or three AI models, compare the indicator sets they propose, and use Ethics Matrices A, B, and C to evaluate and prioritize the results. The AI models will generate possibilities. You remain responsible for deciding which indicators are suitable, how they should be defined, and why they belong in the final project.
Central Question: How can ethics matrices and generative AI be used together to select sustainability indicators without surrendering human judgment to the AI system?
Learning Objectives
- Explain why the selection of sustainability indicators is an ethical and methodological choice.
- Use Ethics Matrices A, B, and C as criteria for generating, screening, and prioritizing indicators.
- Construct a project brief and common prompt suitable for comparison across AI models.
- Compare areas of convergence, divergence, omission, and ambiguity in outputs from two or three AI systems.
- Distinguish a broad sustainability topic from an operational indicator that can be observed, measured, or assessed.
- Select and justify a balanced portfolio of sustainability indicators for the final project.
- Document AI use and retain responsibility for verification, interpretation, and final judgment.
Using the Ethics Matrices as Selection Criteria
The matrices are not being used to produce a mechanical score. Each matrix provides a different set of questions for evaluating proposed indicators.
Matrix | Primary Role in Indicator Selection | Guiding Question |
|---|---|---|
| Ethics Matrix A | Integrity, evidence, transparency, validity, uncertainty, and responsible communication | Can the indicator be defined, supported, measured, and interpreted with integrity? |
| Ethics Matrix B | Broader impacts, public policy, justice, transformation, risk, and precaution | Does the indicator make consequential benefits, burdens, risks, and stakeholder effects visible? |
| Ethics Matrix C | Embedded assumptions, boundaries, categories, proxies, exclusions, and prioritization | How does the indicator itself construct what counts as sustainability? |
Using Penn State AI Studio
Use Penn State AI Studio for the assignment. Begin separate conversations with two or three available AI models so that the outputs remain independent. Upload Ethics Matrices A, B, and C to each conversation, along with the same project-topic brief and the same initial prompt.
Access: Penn State AI Studio
Guidance: Penn State Best Practices for Using AI Tools
Do not upload confidential, proprietary, personally identifying, or otherwise restricted information. Use a project description that contains only information appropriate for course work.
One-Week Workflow
Step | Student Work | Purpose | Output |
|---|---|---|---|
| 1. Define | Prepare a concise brief describing the final project topic, decision context, boundaries, stakeholders, and known constraints. | Give each model the same grounded context. | Project-topic brief |
| 2. Generate | Upload Matrices A, B, and C and run the same base prompt in two or three separate AI models. | Produce independent candidate indicator sets. | Model outputs |
| 3. Compare | Identify convergence, divergence, omissions, weak definitions, unsupported claims, and differences in prioritization. | Evaluate the models rather than accepting one response. | Comparison table |
| 4. Select | Use the matrices to refine and prioritize a final portfolio of indicators. | Make the final ethical and methodological judgments. | Final SI portfolio |
| 5. Reflect | Explain how the matrices changed the AI-generated suggestions and which decisions could not be delegated. | Document responsible AI use and human accountability. | Comparative reflection |
From Sustainability Topics to Operational Indicators
AI systems frequently propose broad themes such as equity, biodiversity, community well-being, affordability, or resilience. These are important domains, but they are not yet operational indicators.
A usable final indicator should specify:
- a clear indicator name and operational definition;
- a unit, calculation, scale, threshold, or assessment method;
- the geographic and temporal boundary;
- the relevant stakeholder, system function, benefit, burden, or risk;
- the connection to one or more ethics matrices;
- a plausible source of evidence or data;
- and an important limitation or possible source of misinterpretation.
The final portfolio should be manageable, sufficiently diverse for the project, and free of unnecessary duplication. The portfolio should also be evaluated as a whole. Several individually useful indicators can still produce an ethically narrow set if they consistently omit a stakeholder group, timescale, or category of impact.
Principles for the AI Comparison
- Use the same project brief and initial prompt with each model.
- Run the models in separate conversations so one model's output does not influence another model.
- Treat model agreement as convergence, not proof that an indicator is correct.
- Treat model disagreement as information about ambiguity, framing, or differing interpretations.
- Verify factual claims, measurement methods, and proposed data sources before using them.
- Do not treat AI-generated citations as reliable until the sources have been independently confirmed.
- Document consequential follow-up prompts and revisions.
- Retain responsibility for the final indicator definitions and priorities.
Assignment 7
Assignment 7: AI-Assisted Ethical Selection and Comparison of Sustainability Indicators
For Assignment 7, you will use two or three AI models to propose sustainability indicators for your final project topic. You will compare the outputs, use Ethics Matrices A, B, and C to evaluate them, and select a final portfolio of six to eight indicators.
The submission will include:
- the common project brief and base prompt;
- a comparison of the AI-generated indicator sets;
- the final portfolio of six to eight operational sustainability indicators;
- a concise explanation of how each selected indicator relates to the ethics matrices;
- and a comparative reflection on model performance, omissions, rejected indicators, verification, and human judgment.
Complete instructions, formatting requirements, and submission links are provided in Canvas.
Readings and Materials
Lesson 7 does not introduce a new survey of sustainability-indicator literature. Return to the bioenergy sustainability-indicator material from Lesson 4 and use that case as a reminder that indicator sets are constructed for particular systems, decisions, stakeholders, and definitions of sustainability.
- Review the sustainability-indicator material assigned in Lesson 4.
- Review Ethics Matrices A, B, and C and the course guides for using them.
- Review the Lesson 7 page on indicator selection and AI comparison.
- Consult Penn State's AI Studio and best-practices guidance as needed.
Main Point
Generative AI can expand the range of possible indicators and make the first stage of indicator development more manageable. AI cannot determine which values should govern a project, whose interests should receive priority, which tradeoffs are acceptable, or what counts as an adequate representation of sustainability. The ethics matrices provide criteria for examining those questions, while comparison across models makes the limits and variability of AI-generated analysis visible.