Generative AI

In the digital realm of text and code,
Where words and algorithms unfold,
There lies a new creation, we're told,
Of language, synthesized and bold.

AI text, born of machine and mind,
With endless knowledge, it can find,
A wealth of information, combined,
Into a language, both clear and refined.

It generates words with ease and speed,
Creating new ideas, to exceed,
The limits of what we once believed,
Was possible, for language to achieve.

Yet, with this power, comes great responsibility,
To use AI text, with care and humility,
For though it brings new opportunities,
We must remember its limitations and fallibilities.

For language is more than words on a screen,
It's a bridge between minds, that can't be seen,
And while AI text may help us glean,
New insights, it can't replace the human dream.

So let us embrace this new frontier,
With wonder, and also, with fear,
And with each new discovery, let us steer,
Towards a future, where AI and humanity, can cohere.

-ChatGPT

 

The use of AI as a tool, both positive and negative, has grown in the recent months. This growth of AI has lead to many questions and fears in Higher Education and other fields. This page along withe the Armacost Library page on Generative AI offers some resources in relation to the growth of Generative AI. The page will also provide information on how professors may work with and around the growing tool while maintaining high standards and a student-focused lens.

AI generated FAQ

The bellow FAQ are generated using an AI tool called ChatGPT

What is AI text?

AI (Artificial Intelligence) is a field of computer science that focuses on creating intelligent machines that can perform tasks that typically require human-level intelligence, such as reasoning, learning, perception, and natural language understanding. In recent years, AI has made significant progress in natural language processing, which is the ability of computers to understand, analyze, and generate human language.

As a college professor, you can use AI text in a variety of ways to enhance your teaching and research. One common use case is for automated grading of assignments and essays. AI text can analyze the structure, grammar, and coherence of student writing and provide feedback and grades accordingly. This can save you a significant amount of time and effort in grading and provide more consistent and objective feedback to students.

AI text can also be used for generating educational content, such as textbooks, study guides, and quizzes. By training AI models on large datasets of educational materials, the system can generate new content that is tailored to specific learning goals and student needs. This can help make educational resources more accessible and personalized for students.

In addition, AI text can be used for research in a variety of fields. For example, in the social sciences, AI text can be used to analyze large datasets of written documents, such as social media posts or news articles, to identify trends, sentiment, and other insights. In the humanities, AI text can be used to analyze and generate new works of literature or to aid in language translation.

However, it's important to note that AI text is not without its limitations and potential drawbacks. For example, automated grading may not capture the full range of qualities that make up good writing, such as creativity, originality, and critical thinking. Additionally, AI text may perpetuate biases and stereotypes that exist in the training data, so it's important to carefully evaluate and monitor the results.

Overall, AI text is a rapidly advancing field with a lot of potential for college professors to enhance their teaching and research. By understanding the capabilities and limitations of AI text, professors can make informed decisions about how to best integrate this technology into their work.

How can I prevent students from using AI text for an assignment ?

Preventing students from using AI text for an assignment can be challenging, as there are many different tools and services available that offer AI-generated text. However, there are a few strategies you can consider to discourage or prevent the use of these tools:

  1. Provide clear guidelines and expectations: Be clear with your students about what is expected of them for the assignment, and emphasize the importance of original work. Clearly state your policy on the use of AI text and make it clear that any use of these tools will be considered a violation of academic integrity.

  2. Use plagiarism detection software: Consider using plagiarism detection software, such as Turnitin or PlagScan, to check for instances of AI-generated text. These tools can help identify text that has been copied or pasted from online sources or generated by AI, and can help deter students from attempting to use these tools.

  3. Use open-ended assignments: Consider assigning open-ended tasks that require critical thinking, analysis, and interpretation. These types of assignments are less likely to be susceptible to AI-generated text, as they require original thought and reasoning.

  4. Monitor student work: Keep a close eye on student progress throughout the assignment, and be on the lookout for signs of AI-generated text, such as inconsistent writing styles or grammar errors. You can also consider conducting oral exams or in-person interviews to assess students' understanding of the material and identify any discrepancies between their written work and their knowledge.

  5. Engage with students on academic integrity: Finally, make sure to discuss the importance of academic integrity and the consequences of violating it. Help students understand the ethical implications of using AI-generated text and emphasize the importance of developing critical thinking skills and producing original work.

Where does AI text get its information?

AI text can get its information from a variety of sources, depending on the specific algorithm and application. Some common sources of information for AI text include:

  1. Training data: AI text models are often trained on large datasets of text, such as books, articles, or social media posts. The model learns patterns and structures in the text, and uses this knowledge to generate new text.

  2. Knowledge bases: AI text can also be built on top of knowledge bases, which are databases of structured information about a particular domain. For example, a knowledge base of medical information might include information about diseases, symptoms, treatments, and medications. AI text can use this information to generate descriptions, explanations, or instructions.

  3. Search engines: Some AI text applications use search engines to gather information from the web. The system can extract relevant information from web pages, such as descriptions, definitions, or statistics, and use this information to generate new text.

  4. User input: In some cases, AI text may rely on user input to generate new text. For example, a chatbot might use natural language processing to understand a user's request or question, and generate a response based on that input.

It's important to note that the quality and accuracy of AI-generated text depends on the quality and relevance of the underlying information. If the training data or knowledge base is incomplete or biased, for example, the AI text may also be incomplete or biased. Additionally, AI text may not always be able to capture the nuances and complexities of human language, and may sometimes produce errors or inaccuracies.

How can students benefit from AI text?

Students can benefit from AI text in a variety of ways. Here are some examples:

  1. Writing assistance: AI text can help students improve their writing by providing feedback on grammar, style, and coherence. For example, some AI tools can suggest better word choices or highlight areas where the writing is unclear or confusing. This can help students refine their writing skills and produce more polished and effective work.

  2. Research assistance: AI text can also help students conduct research more efficiently. For example, some AI tools can analyze large datasets of academic articles and identify relevant information and connections between them. This can save students time and help them find more relevant and useful sources of information.

  3. Language learning: AI text can be useful for language learners by providing interactive exercises and real-time feedback on pronunciation, grammar, and vocabulary. Some AI language learning tools can even generate conversation prompts and responses, providing students with an opportunity to practice their language skills in a realistic context.

  4. Accessibility: AI text can also be beneficial for students with disabilities, such as visual impairments or dyslexia. Text-to-speech software, for example, can read text aloud, making it easier for students with visual impairments to access written materials. Similarly, AI-powered text recognition software can help students with dyslexia read and understand text more easily.

  5. Personalization: AI text can be tailored to individual student needs and learning styles. By analyzing data on student performance and preferences, AI tools can provide personalized feedback, recommendations, and study materials. This can help students learn more effectively and efficiently by focusing on their individual strengths and weaknesses.

Overall, AI text has the potential to support and enhance student learning in a variety of ways. By providing personalized feedback, improving writing skills, and helping with research, language learning, and accessibility, AI text can help students succeed and achieve their academic goals.

How can students be harmed by AI text?

While AI text can be beneficial for students in many ways, there are also potential harms and risks associated with its use. Here are some examples:

  1. Ethical concerns: There are ethical concerns around the use of AI text, particularly in cases where the technology is used to generate fake news, propaganda, or hate speech. This can be harmful to students, as it can lead to a lack of trust in information sources and contribute to the spread of misinformation and disinformation.

  2. Plagiarism: AI text can make it easier for students to plagiarize by generating text that is very similar to existing sources. This can undermine academic integrity and lead to disciplinary action or sanctions.

  3. Overreliance on technology: If students rely too heavily on AI text to do their work, they may miss out on opportunities to develop critical thinking, research, and writing skills. This can hinder their ability to learn and succeed in future academic and professional settings.

  4. Inaccuracy: AI text may not always be accurate or reliable, particularly if the underlying data or algorithms are biased or flawed. If students rely on inaccurate information generated by AI text, they may produce incorrect or misleading work, which can be harmful to their academic standing or future career prospects.

  5. Reduced human interaction: If students use AI text for language learning, they may miss out on opportunities to practice their language skills in real-life contexts with other speakers. This can limit their ability to communicate effectively in the language and hinder their future language learning and career opportunities.

Overall, it's important for students to use AI text with caution and to be aware of its potential risks and limitations. By balancing the benefits and harms of AI text, students can make informed decisions about how to use this technology to support their learning and achieve their academic goals.

How can I integrate AI text with my pedagogy?

Integrating AI text into your pedagogy can help enhance student learning and engagement. Here are some tips on how to do this:

  1. Use AI text to provide feedback: AI text can provide immediate feedback to students on their writing, including grammar, style, and coherence. By incorporating AI text tools in your feedback process, you can help students improve their writing skills and produce more polished and effective work.

  2. Incorporate AI text in research assignments: AI text can help students conduct research more efficiently by analyzing large datasets of academic articles and identifying relevant information and connections between them. By assigning research assignments that incorporate AI text tools, you can help students find more relevant and useful sources of information.

  3. Use AI text for language learning: AI text can be useful for language learners by providing interactive exercises and real-time feedback on pronunciation, grammar, and vocabulary. By incorporating AI language learning tools in your language classes, you can provide students with an opportunity to practice their language skills in a realistic context.

  4. Assign assignments to evaluate and critique AI text: By asking students to evaluate and critique AI text and the algorithms behind them, you can encourage critical thinking and reflection about the ethical implications of AI text technology.

  5. Encourage active learning and engagement with AI text: To ensure that students are engaged and actively learning when using AI text, consider designing assignments that require students to actively interact with the AI tool or incorporate the AI text into class discussions and debates.

Overall, by integrating AI text into your pedagogy, you can help students develop critical thinking skills, improve their writing and language skills, and engage more deeply with the material. As with any new technology, it's important to consider the potential benefits and risks of AI text and to use it in a responsible and effective way that enhances, rather than replaces, traditional teaching practices.

Recommended strategies for Generative AI in your class

Note: Instructors should think about Universal Design Learning when working on solutions to AI texts.

Re-evaluate you assignments

ChatGPT and AI Texts produce well written text gathered, but is not personalized. To reduce the use of AI text tools by students it is helpful to re-evaluate your assignments and ask:

  1. Is the prompt generic? Generic prompts that ask students to simply define an item or list chronological events can easily be answered by ChatGPT and have little effect on critical thinking. Try expanding prompts that incorporate critical thinking or uniqueness in the response.
  2. Is the assignment student centered? Students will want to write about something they are interested in. If the assignment has little connection to the student or it is not clear why the student should care about the assignment they may look for an easier solution. Try providing enough openness to the prompt so that students can focus on topics they care about or, in the assignment, explain the goals of the assignment.
  3. What are the goals of the assignment? Review the goals you had when creating the assignment, could students (or yourself) view these goals as tedious or busy work? If so, change the assignment to have more robust goals or be clear on the usefulness of the goals. Finally, if in reviewing the goals you find the assignment does not actually promote them, change the assignment to better align with the goals.
  4. Is writing the best way to reach the goal? Review if other ways for students to show learning could be used. AI text is hard to use if the final product is a video of the student or a in class presentation.

Discuss with students the value of writing

Students may come into your class being trained to think that writing is a formulaic practice that just regurgitates information. Discussing with the students on the utility of writing, how it can be used to not only convey information but formulate it, can lead students to reassess writing assignments.

Discuss with students AI text

Being open with the students about why AI text is not appropriate for this assignment not only shows respect for the students but also builds digital literacy.

  1. AI text can incorporate false information – AI text generators use previously created material to form their answers. Their writing may be structurally correct, but the generators are not sophisticated enough to evaluate the information used to create the material. This leads to inaccurate or false information to be present.
  2. AI text can be biased – AI tools have had a history of race and gender bias; AI text is no different. Depending on the material used to create the answer, the biases of the material can be amplified.
  3. AI text can lead to homogenization – Expanded use of the AI text may be mechanically or grammatically correct but reduces the natural mistakes or creative writing found in human created text. This can reduce the growth of the language or incorporation of other cultures.
  4. Academic honesty and AI – Talking with the students about academic honestly, why it is important, and why AI text may violate academic honesty provides students the opportunity to critically understand policies.

Incorporate AI tools

AI text is not going away, you may find it interesting and powerful to have students use it in a course.

  1. Have the students ask an AI text tool a question then have them review and analyze the answer. This will build their digital literacy as they will be able to find trends in AI text, compare the answers to the knowledge presented in class, and grow a critical understanding of the subject. Students may also compare the responses of two different AI tools.
  2. Use AI tools for basic research – have students do initial research using the AI tools. Responses from AI may help direct the students in an assignment. NOTE – ChatGPT does not provide citations so students will not be able to follow up information produced.

CAS Sample Policies

Below are some sample policy text recommended by CAS. This text can be added to your syllabus so that the students know the extent AI generated material can be used in the course.

  • Zero tolerance: In this course, there is no ethical use for generative AI tools such as ChatGPT. All assignments ask you to use only resources from our course or other approved sources to develop your ideas at all stages of writing. Use of generative AI at any stage of a project, including outlines and drafts, will be considered a violation of the university’s academic honesty policy. 
  • Case by case: In this course, we will discuss how to use AI ethically. For each assignment, we’ll talk about how generative AI might be used, how to cite it, risks of using it for that project, and what contributions and ideas you will be personally responsible for. From outlines, to drafts, to feedback on drafts, we’ll consider how we want to use generative AI and how we want to integrate our own ideas, course discussions and questions, course readings, and other approved sources. Using generative AI in ways outside what we agree to as a class for any assignment will be considered a violation of the university’s academic honesty policy. 
  • All the AI that is useful: In this course, we will discuss how to use AI ethically. I will assume you are using AI to the extent it might be useful for every project and that you will cite it accordingly, just as you would any other source. Grades and assessment will address knowledge, skills, and expertise you need to meet course learning objectives with generative AI as a resource, just as you might use Google, meetings with an Armacost librarian, or time with me in office hours. Failure to cite generative AI use will be considered a violation of the university’s academic honesty policy. 
  •  AI Unlimited: In this course, you are invited and expected to use AI to the extent it might be useful for every project. There is no need to cite AI in this course. Please note, your other courses may have different expectations for use of AI, including not using it at all. Failure to follow guidelines for generative AI use for any course will be considered a violation of the university’s academic honesty policy.  

Citation

APA - How to cite ChatGPT,  Timothy McAdoo April 7, 2023,  APA Style Guide Blog

Chicago - Documentation of Sources FAQ,  The Chicago Manual of Style Online

MLA - How do I cite generative AI in MLA style?, Ask The MLA

External Articles

Bellow are external articles about AI text and art

General

Guidance for generative AI in education and research - UNESCO 2023. - UNESCO's recomendations and guidance for ethical AI use in education and research.

Text

Challenging systematic prejudices: an investigation into bias against women and girls in large language models - UNESCO 2024 - A UNESCO study looking at large language model AI output and their relation to gender and sexuality bias.

8 Ways to Create AI-Proof Writing Prompts, Erck Ofgang 3/17/2024 Teach and Learning.

AI detectors discriminate against non-native speakers, says stanford research, Erick Ofgang, 11/8/2023 Teaching And Learning - Link to the research paper.

Unexpected Bedfellows: Using ChatGPT to Uphold Academic Assessment Integrity, Arno Louw, 9/20/2023, Educause - An articial about how you can use ChatGPT to assist you in assess students work. 

Deconstructing ChatGPT on the Future of Continuing Education, Ray Schroeder 12/14/2022, Inside Higher Ed- An article breaking down an AI-generated poem, discussing the growth of the technology.

ChatGPT Both Is and Is Not Like a Calculator, Jon Warner 1/16/2023, Inside Higher Ed - Warner contemplates the best way to manage AI text in the classroom.

Freaking Out About ChatGPT—Part I, Jon Warner 1/5/2022, Inside Higher Ed - Warner talks about the worries about ChatGPT, their experience with imputing an AP exam prompt, and reflection on why worries of ChatGPT make visible issues with some writing assignments/assessments.

Outsmart ChatGPT: 8 Tips for Creating Assignments It Can’t Do, Alyson Klein 2/14/2023,  Education Week - Tips about designing assignments that reduces the students' ability of using AI text.

Chronicle of Higher Education AI search, Chronicle of Higher Education - Link to the Chronicle's search results for AI. (Free Access with university account)

Art

AI Generated Art for a Comic Book. Human Artists Are Having a Fit, James Hookway 1/29/2023, The Wall Street Journal - An article discussing how AI art has led to copyright questions and ethical concerns

Lawsuits over Stability AI's Stable Diffusion could threaten the future of AI-generated art, Asia Martin 1/25/2023, Insider - Discusses the copyright issues that have arrived with AI art, focusing on Stability AI.

AI art generators face separate copyright lawsuits from Getty Images and a group of artists, Beatrice Nolan 1/19/2023, Insider - Lawsuits from Getty images and three artists against AI art generators.

Other Higher Education Institution's pages on AI

James Madison University- ChatGPT by Emily Gravett 1/26/2023

UC Berkeley - Understanding AI Writing Tools and their Uses for Teach and Learning at UC Berkeley.

University of Toronto - ChatGPT and Generative AI in the Classroom

Other cultivated resources

AI Text Generators: Sources to Stimulate Discussion among Teachers - Google Doc complied by Anna Mills for the Writing Across the Curriculum Clearing house.
 
Exploring AI Pedagogy: A Community Collection of Teaching Reflections - A collection of past classroom AI activities. The collection is part of the MLA-CCCC Joint Task Force on AI and writing