What are the best practices around AI use at Tomorrow University?

Guidelines on leveraging AI throughout your academic journey at Tomorrow University

Tomorrow University exists to support changemakers. All of the assessments and more generally the challenges that you as a learner experience at ToU are designed to power up your capacity to create change through your own projects and initiatives.

As part of our commitment to promoting innovation and collaboration within the fields of sustainability, entrepreneurship, and technology, we encourage you to explore the use of Artificial Intelligence (AI) tools in your student assignments. These tools offer a wealth of opportunities for learning and creativity, enabling you to leverage existing technologies while contributing to the academic community.

However, it is essential to balance the advantages of AI generated outputs with the responsibility to avoid plagiarism and give proper credit to original creators. To help you navigate this exciting landscape, we have outlined both rules of use and guidelines for using AI in your ToU assignments:

Rules of use The following must be adhered to when completing challenges with the assistance of AI:

1. Attribute Properly: When incorporating generative AI model outputs into your assignments, give appropriate attribution to the original creators. Provide citations and references to the relevant repositories, research papers, or documentation. This acknowledgment demonstrates respect for the intellectual property of others and supports a culture of sharing and collaboration. Copying texts of third parties in whole or in part, verbatim or almost verbatim, and passing them off as your own work is considered plagiarism, which is a form of academic misconduct and can lead to serious consequences such as being marked with a failing grade or failing a whole course. In severe cases of plagiarism, even criminal charges can be filed. For more information, see the ToU plagiarism policy.
2. Document Your Process: When using generative AI models to help with your assignments, keep a detailed record of the generative AI model outputs you used, along with the modifications you made and how you integrated them into your assignment. Make sure you clearly differentiate between the work of the model and your own competency-enhancing work, by stating how the model was used.

For example - “Text based language models where used to generate the first outline of this written assignment - which was then improved by independent research and writing of the main text” or “Generative models for image were used to create the figures in this presentation” etc 3. Follow the AI use terms stipulated by the challenge owner: Each challenge has different rules on the use of AI: Prohibited, permitted, partially permitted. These will be communicated by the challenge owner, and must be adhered too (see challenge specific AI usage below).

Guidelines The following is a guide to AI use best practice

1. Consider the Competencies:

   When a particular Challenge is designed to support the development of a particular competency that you would no longer get to practice by using AI to create your work, choose to produce your own work instead and/or significantly modify AI outputs based on your own work. For example, if presentation and storytelling skills are key competencies in a Challenge, your development will be best realized by completing your work without AI-generated presentations. In this case, AI outputs should supplement your own work, rather than replace it or give feedback on your work to improve it.

2. Choose Reputable AI Platforms: When utilizing AI tools, opt for reputable platforms and libraries that have active communities and regular updates. Popular platforms such as TensorFlow, PyTorch, and scikit-learn offer a vast range of functionalities and resources for your AI-related tasks.

3. Attribute Properly: See rule 1 above.

4. Understand and Adapt Open Source Code: When using AI generated code, or code from open source repositories, make sure you fully understand the underlying algorithms and methodologies. Take the time to review and adapt the code to suit your specific assignment requirements. Always credit the original source and maintain a clear distinction between your contributions and the open source elements you used.

5. Check AI sources When using AI for research, it is your responsibility to verify the accuracy of the sources provided. Some AI models “hallucinate”, and provide non existent sources, or match poor sources to a particular piece of information. To avoid this, use AI models which specifically provide research support (Elicit, Semantic scholar, supper symmetry, consensus) rather than general text-based models, and double check all sources obtained by AI.

6. Check for License Terms: Review the licensing terms of the open source AI tools you use. Different licenses come with specific requirements, such as crediting the original authors or sharing your modifications under the same license. Comply with the terms of the licenses to ensure ethical use and proper sharing of your work.

7. Do not use confidential information for input: If challenge owners / partners provide confidential information for a challenge, do not use this information directly for AI prompts as this violates data protection policies.

8. Document Your Process: See rule 2 above.

9. Share Your Contributions: If you make improvements, modifications, or extensions to open source AI code as part of your assignment, consider sharing your work with the community. Contributing back to the open source projects you used can be a rewarding experience and allows others to benefit from your innovations.

10. Seek Guidance: If you have any uncertainties about using open source AI in your assignments or concerns about plagiarism, don't hesitate to seek guidance from your Challenge Owners or supervisors. They can provide valuable insights and ensure that your work aligns with the university's academic integrity standards.

Embracing open source AI in your assignments can foster a culture of collaboration, innovation, and knowledge sharing within our university community. By adhering to these guidelines and being mindful of plagiarism, you will contribute positively to the advancement of sustainable practices, entrepreneurial endeavours, and technological solutions.

Remember, academic integrity is of the utmost importance, and giving proper credit to the source is not only a requirement but also an essential part of the learning process.

Guidelines based on assessment types

Reports:

Written work examining a specific topic (usually including recommendations), requiring research, analysis, critical evaluation and development of convincing/well supported arguments

• Use AI to supplement your ideas for talking points and general structure. Do not use AI to develop the core idea of your work.
• You may use AI to help develop thesis/essay text using your ideas as prompts. However, make sure to adjust this text based on your own research and understanding of the topic. Do not copy and paste large blocks of text from AI models without revision.
• Do not solely rely on AI to develop recommendations from the research. These should supplement your own critical evaluation of the literature, data or evidence.
• When conducting research, please use AI specifically meant to support research (such as Elicit, Semantic scholar, supper symmetry, consensus etc) rather than general text-based models.
• Review all AI citations, and do not rely solely on these. Engage directly with the research in order to gain an understanding of the topic’s context. Additionally, make sure all citations and resources are up to date and relevant (beware of AI “hallucinating” sources).
• Clearly attribute any text or information obtained from AI tools.

Conclusion: Reports are useful for developing your own ideas, research, analysis and evaluation skills, so avoid using AI in place of developing these competencies.

Thesis/essays:

Written work consisting of a written review and evaluation of the literature. Requires research, analysis, critical evaluation and development of convincing/well supported arguments

• Use AI to supplement your ideas for talking points and general structure. Do not use AI to develop the core idea of your work.
• You may use AI to help develop thesis/essay text using your ideas as prompts. However, make sure to adjust this text based on your own research and understanding of the topic. Do not copy and paste large blocks of text from AI models without revision.
• When conducting research, please use AI specifically meant to support research (such as Elicit, Semantic scholar, supper symmetry, consensus etc) rather than general text-based models.
• Review all AI citations, and do not rely solely on these. Engage directly with the research in order to gain an understanding of the topic’s context.
• Clearly attribute any text or information obtained from AI tools.

Conclusion: A thesis or essay is useful for developing your own research and review skills. Including ideation, analysis and evaluation. Avoid using AI in place of developing these competencies.

Presentations:

Presentations (video or in person) communicate your ideas (research and/or proposals) to an audience using slides, multimedia, and oral delivery.

• The major content of the presentation should consist of your own work (see comments on other assessment types above).
• Use AI to create a general outline/structure and suggest talking points for the presentation using your previous research as prompts. Use your own research and experience to modify this outline, and do not use out of the box outlines.
• You may use AI to create presentation slides and images using your own knowledge and research as prompts.
• You may use AI to assist in the development of your presentation script, but avoid developing the entire script with AI tools. Additionally, do not use AI speech for your presentation/video recording. Presenting your own work in a clear and engaging manner is a major competency for this type of assessment.
• If using open AI at any point in your presentation, clearly explain your process and properly accredit sections which utilized AI. Transparency around any AI augmentation enables proper assessment of your skills.

Conclusion: Presentations are not only useful for developing your creative design and public speaking skills, but also for demonstrating your subject knowledge and critical evaluation skills. AI tools can be used to supplement ideas, but not in place of developing these competencies

Coding challenges

Coding challenges: Develop your coding language skills in order to complete tasks / problems

• Do not copy and paste entire code segments from AI models without review or modification. Understanding your code is a major competency for this type of assessment. Not only does this hinder your learning but is likely to result in errors.
• Use AI to help you understand unfamiliar algorithms or code.
• If using open AI for the development of coding challenges, clearly explain your process and properly accredit coding sections generated using these tools. Transparency around any AI augmentation enables proper assessment of your skills.

Conclusion: Coding challenges are not only useful for developing your coding skills but also for demonstrating your problem solving skills. AI tools can be used to supplement these skills, but not in place of developing these competencies