Overview: DesignPLUS

Artificial Intelligence (AI) is the simulation of human intelligence in machines that are programmed to think and learn like humans. AI aims to enable machines to perform tasks that typically require human intelligence, such as recognizing speech, making decisions, solving problems, and understanding natural language. This technology encompasses various subfields, including machine learning, where algorithms improve automatically through experience; natural language processing, which enables machines to understand and interact using human language; and computer vision, where AI interprets and understands the visual world. AI has the potential to transform numerous industries and disciplines, including education, healthcare, finance, and more, by enhancing efficiency, personalizing experiences, and unlocking new possibilities through its adaptive and predictive capabilities. While there are numerous ways in which AI can be implemented in the classroom, from Individual Tutoring Systems to Augmented and Virtual Reality, this resource is dedicated to the use of -generative AI in the classroom such as ChatGPT pr MidJourney.

Artificial Intelligence

Definitions

Vocabulary of Common AI Terms

AI Bias

AI bias refers to systematic and unfair disparities that can arise in artificial intelligence systems, stemming from biased data, algorithms, or the subjective decisions of those involved in AI development. This bias can manifest in various forms, such as racial, gender, or socioeconomic bias, leading to skewed or discriminatory outcomes that reinforce existing inequalities. For instance, if an AI model is trained on historical hiring data that contains gender bias, the system might undervalue applications from women. The challenge of AI bias is significant because AI technologies are increasingly used in critical decision-making processes, from job recruitment to loan approvals and legal sentencing. Addressing AI bias involves a multifaceted approach, including diversifying training data, implementing ethical guidelines in AI development, and continuously monitoring and adjusting AI systems to ensure fairness and equity. The goal is to create AI systems that are not only intelligent and efficient but also unbiased and equitable, promoting fairness and inclusivity in the digital age.

Shedding light on AI bias with real world examples

What Do We Do About the Biases in AI?

There's More to AI Bias Than Biased Data

Algorithm

Algorithms are defined as finite sets of precise instructions or rules for performing a task or solving a problem. They are the foundational concepts not only in computer science and artificial intelligence but also in mathematics and other disciplines where systematic procedures are required for problem-solving. An algorithm provides a step-by-step procedure for calculations, data processing, automated reasoning, and other tasks, ensuring that if followed correctly, a desired output will be produced from a given input.

In the context of artificial intelligence, algorithms are crucial for enabling machines to learn from data, make decisions, and predict outcomes. They range from simple, rule-based algorithms for straightforward tasks to complex algorithms used in machine learning and deep learning, capable of handling vast amounts of data and making sophisticated decisions.

Additional Resources

Understanding algorithms in Computer Science

Harvard Professor Explains Algorithms in 5 Levels of Difficulty

Introduction to Algorithms

Computer Vision

Computer Vision is a field within artificial intelligence and computer science that focuses on enabling machines to interpret, analyze, and make decisions based on visual data from the physical world. This discipline involves the development of algorithms and systems that can process, analyze, and understand images or multidimensional data to perform tasks typically requiring human visual comprehension. Computer vision aims to replicate and exceed human visual abilities in machines, allowing them to recognize objects, faces, scenes, and activities in images and videos, often in real time.

Additional Resources

What is computer vision?

Everything you wanted to know about computer vision

Computer Vision: Crash Course Computer Science

Data Mining

Data Mining is a critical process in the field of Artificial Intelligence (AI) that involves the extraction of valuable patterns, correlations, and insights from large and complex datasets. This process employs a blend of statistical, machine learning, and computational techniques to sift through vast amounts of data from various sources, such as databases, websites, and other repositories, identifying hidden patterns and unknown correlations. In the context of AI, data mining serves as a foundational step that enables intelligent systems to learn from historical data, make informed predictions, and improve decision-making processes. By uncovering underlying structures and relationships within the data, data mining facilitates the development of models that can anticipate future trends, classify data into meaningful categories, and support the automation of complex tasks.

Additional Resources

What is Data Mining and Why is it Important?

What is Data Mining? A Beginners Guide

Definitions, Benefits, Applications and More

Deep Learning

Deep Learning is an advanced subset of machine learning characterized by its use of deep neural networks, which are essentially algorithms modeled loosely after the human brain's structure and function. These networks consist of multiple layers of interconnected nodes or "neurons," each layer designed to process different aspects of the input data and pass on the processed information to subsequent layers. The "depth" in deep learning refers to the number of these layers, with more layers allowing for the processing of increasingly abstract and complex features. This structure enables deep learning models to automatically learn and identify intricate patterns in vast datasets, improving their performance in tasks such as image and speech recognition, natural language processing, and decision-making as more data is fed into them. Deep learning has revolutionized fields like computer vision and AI by enabling machines to perform tasks with an accuracy that approaches or even surpasses human capabilities in some domains.

Additional Resources

What is Deep Learning

A Deep Learning Tutorial for Beginners

Deep Learning vs Machine Learning

Generative Pretrained Transformer (GPT)

The Generative Pretrained Transformer (GPT) is an advanced type of artificial intelligence model known for its ability to generate coherent and contextually relevant text based on the input it receives. Developed by OpenAI, GPT models are built on a transformer architecture—a deep learning model that relies on mechanisms called attention and self-attention to weigh the importance of different words in a sentence, regardless of their positions. These models are "pretrained" on vast datasets compiled from the internet, allowing them to learn a broad range of language patterns and knowledge before being fine-tuned on more specific tasks. This pretraining process enables the models to generate text that is not only fluent and grammatically correct but also contextually nuanced. GPT models are versatile and can be adapted for a variety of applications, including conversation agents, text summarization, translation, and even creative writing.

Transformers Explained

Large Language Models

Large Language Models (LLMs) are sophisticated artificial intelligence systems specifically designed to process, generate, and interact using human language. These models are built on transformer architectures and are trained on vast amounts of text data, allowing them to grasp the subtleties and complexities of natural language. LLMs excel in a range of linguistic tasks, including text generation, translation, summarization, and engaging in conversations, by predicting subsequent words based on prior context. Their training involves extensive adjustment of parameters through supervised learning, which helps the models learn from example inputs and outputs. The enormous scale of these models, often featuring billions or trillions of parameters, enables them to capture a wide spectrum of linguistic patterns and information, enhancing their versatility across various domains.

Additional Resources

Large Language Models Explained

Large language models can do jaw-dropping things. But nobody knows exactly why.

What is retrieval augmented generation?

Natural Language Processing

Natural Language Processing (NLP) is a pivotal area of Artificial Intelligence (AI) focused on the interaction between computers and humans through natural language. The goal of NLP is to enable computers to understand, interpret, and generate human language in a way that is both valuable and meaningful. This involves a complex blend of computer science, artificial intelligence, and linguistics, aimed at bridging the gap between human communication and machine understanding. NLP encompasses a range of tasks, including speech recognition, sentiment analysis, language translation, and chatbot development. Through these capabilities, NLP allows machines to perform tasks such as summarizing texts, answering questions, and even engaging in conversational dialogue with users. The advancement of NLP technologies has significant implications for various fields, enhancing the accessibility of information, automating customer service, and opening new avenues for human-computer interaction.

Additional Resources

What is Natural Language Processing?

The Power of NLP

The role of natural language processing in AI

Supervised Learning

Supervised learning is a cornerstone methodology in the realm of machine learning and artificial intelligence, where models are trained on a labeled dataset. This approach involves providing the algorithm with input data paired with the correct output, enabling the model to learn by example. The primary goal of supervised learning is to construct a predictive model that can make accurate predictions or decisions when faced with new, unseen data by identifying patterns and relationships in the training data. Common applications of supervised learning include classification tasks, where the model categorizes input data into predefined classes, and regression tasks, where it predicts a continuous value. The effectiveness of a supervised learning model is heavily dependent on the quality and comprehensiveness of the training data.

Additional Resources

Tech Target: Supervised Learning

Supervised Machine Learning

Supervised vs Unsupervised vs Reinforcement Learning

Unsupervised Learning

Unsupervised learning is a type of machine learning where algorithms are tasked with identifying patterns and structures in data without any external guidance or labeled outcomes. This form of learning is crucial when the data lacks annotations or when exploring unknown relationships within the dataset. In unsupervised learning, the model attempts to organize the data, typically through clustering or dimensionality reduction techniques, to discover the intrinsic groupings and connections. Clustering algorithms, for example, segregate datasets into distinct groups based on similar characteristics without prior knowledge of group definitions, while dimensionality reduction techniques reduce the number of variables under consideration, simplifying the model while retaining essential information. Unsupervised learning is particularly valuable for exploratory data analysis, anomaly detection, and complex systems modeling where the underlying patterns are not initially evident. It offers a powerful tool for gaining insights from data, allowing researchers and analysts to uncover hidden patterns and make data-driven decisions in the absence of explicit instructions.

Additional Resources

Introduction to Unsupervised Learning

Unsupervised Learning: Algorithms and Examples

Using AI in the Classroom

Using Generative AI in the Classroom

Generative AI tools, like language models and content creators, can assist in generating educational content, crafting personalized learning materials, and providing instant feedback on student assignments. These technologies can also facilitate interactive simulations and scenario-based learning, helping students grasp complex concepts through engaging, adaptive, and immersive experiences. Furthermore, Generative AI can support faculty by automating routine tasks such as grading and exam preparation, allowing more time for direct student interaction and pedagogical refinement. However, incorporating these tools requires careful consideration of ethical implications, such as ensuring fairness, maintaining academic integrity, and safeguarding student data privacy.

How much AI?

Level of AI Integration

One of the initial steps faculty members should consider is gauging the appropriate level of AI integration in their courses. It can be complex to ascertain how students are utilizing AI to assist with their assignments, as well as deciding whether to incorporate AI as a pedagogical tool in the classroom. To facilitate this process, we recommend adopting the Stoplight system. This approach allows you to categorize different aspects of AI use into clear, manageable levels, helping you make informed decisions about integrating technology effectively and responsibly into your teaching strategy.

Red Light

Untitled design.svg The "Red Light" category in the Stoplight system indicates a setting where no AI tools are used within the course. This approach is typically chosen for courses where traditional teaching methods are preferred, or where the integrity and assessment of fundamental skills are prioritized without the assistance of AI. Opting for a Red Light scenario might be particularly relevant in disciplines where personal judgment, critical thinking, or manual processes are essential to the learning objectives. This level ensures that students develop and demonstrate their competencies without reliance on AI support, preserving academic rigor and fostering a direct understanding of the subject matter through conventional instructional methodologies.

Yellow Light

The "Yellow Light" category in the Stoplight provides for student AI use under specific contexts. In this setting, AI is utilized strategically to enhance, reinforce, and clarify course material. Students are encouraged to use AI tools as a supplementary resource to deepen their understanding and support their learning. However, there are stringent rules to instructor control over use. This may include students submitting their prompts and interactions to their instructors for review. This transparency ensures that the use of AI is directly related to learning objectives and not for completing assignments independently. Importantly, directly copying information from AI outputs is strictly prohibited and is considered a serious violation of the academic integrity policy. This approach allows students to benefit from AI technologies while maintaining a strong foundation of personal accountability and academic rigor.

Green Light

The "Green Light" category in the Stoplight system represents the most liberal use of AI by students in the classroom. In this scenario, AI is fully integrated into the learning process, allowing students to use AI tools extensively to explore concepts, generate ideas, and assist in creating drafts for assignments. Students are encouraged to harness AI for a broad range of academic activities, from researching and organizing information to refining their writing and analytical skills. Under the Green Light, there is a significant emphasis on using AI as a collaborative tool that complements the student’s efforts. However, students are still expected to critically engage with and revise AI-generated content, ensuring that the final submission reflects their own understanding and critical analysis. This level promotes a progressive approach to learning, where AI acts as a powerful ally in educational development, fostering innovation and creativity while enhancing learning outcomes.

Example Syllabus Statements

Examples Currently Used by UNK Faculty

Dr. Sri Sheshadri: Marketing/Agribusiness/CSM

“AI/ChatGPT: This course embraces transparency in the integration of AI throughout the curriculum, encompassing its use in assignments and teaching support. Students will document and reflect on their generative AI utilization within their major assignment portfolios. While generative AI will play a role in your writing process, it will not be employed for grading. The primary feedback will come from the instructor, with occasional AI assistance for specific writing strategies or clarification. Unauthorized use entails not documenting, undermining learning, or misrepresenting AI-generated content as original work. We will collectively explore ethical AI use and its impact on your learning. This policy was written with the help of ChatGPT.

Dr. Amanda Sladek: English

AI writing software (such as ChatGPT) is part of the writing landscape. Like grammar check and Google, it is a tool that can help your writing or make it worse. Unlike these tools, though, it raises ethical concerns related to plagiarism, intellectual property, and copyright. Overreliance on AI can also prevent you from learning and practicing the skills this class will teach you. Because it is essential that we learn to use AI responsibly and ethically, you are allowed to use AI writing tools to assist with prewriting (outlining, idea generation, etc.) and some editing (if you have trouble wording a sentence, for instance). However, the ideas you present in this class and the majority of your writing is expected to be your own. Any AI tools you use must be referenced in your Works Cited and explained in your project reflection.

Examples from other institutions

Examples from Other Universities

These statements are available here under Creative Commons License CC BY-NC-SA 4

Ryan Baker: Learning, Teaching, and Literacies Division, University of Pennsylvania

Within this class, you are welcome to use foundation models (ChatGPT, GPT, DALL-E, Stable Diffusion, Midjourney, GitHub Copilot, and anything after) in a totally unrestricted fashion, for any purpose, at no penalty. However, you should note that all large language models still have a tendency to make up incorrect facts and fake citations, code generation models have a tendency to produce inaccurate outputs, and image generation models can occasionally come up with highly offensive products. You will be responsible for any inaccurate, biased, offensive, or otherwise unethical content you submit regardless of whether it originally comes from you or a foundation model. If you use a foundation model, its contribution must be acknowledged in the handin; you will be penalized for using a foundation model without acknowledgement. Having said all these disclaimers, the use of foundation models is encouraged, as it may make it possible for you to submit assignments with higher quality, in less time.

The university's policy on plagiarism still applies to any uncited or improperly cited use of work by other human beings, or submission of work by other human beings as your own.

Christy Snider: History, Barry College

This class allows the use of AI tools (such as ChatGPT, Bing chat, and others) for your out-of-class assignments. AI is a new and valuable skill to master. Students who use AI correctly can become more productive, efficient, and skilled scholars. AI should not be viewed as a way to avoid learning the material, but rather as an assistant that can help you master content and produce better results.

Artificial intelligence is a rapidly evolving field that has many applications and implications for various disciplines. Different courses and instructors will have different policies regarding the use of AI tools and services for academic purposes. It is your responsibility to follow the AI policies for each of the courses that you are enrolled. Violating the AI policies of any course may result in serious consequences, such as a lower grade, a failing mark, or academic probation.

1. To get high quality results from AI, you need to craft good prompts. Simple prompts lead to weak results. OpenAI the company that built ChatGPT has put out documentation on some of the best practices to use with AI chats to achieve good results.

2. Don't believe anything that AI tells you. If it provides a number or a fact, verify it with a trustworthy source. You will be held accountable for any mistakes the tool produces.

AI fabricates or ""hallucinates"" seemingly credible data all the time. It can generate wholly inaccurate content that is nonetheless highly persuasive. This is especially true when asking it for references, quotations, citations, and calculations. Among the various different models, Bing tends to have the least hallucinations, primarily due to its internet connectivity. However, Bing will still lie to you; especially if you ask it to expand upon a topic for which it doesn't have information. AI can also produced biased answers. It was likely trained on material primarily created by by straight, white, men. Your grade will be penalized for each wrong fact or inappropriate text submitted. Make sure, therefore, to closely read and verify anything AI generates to ensure that it accurately reflects the facts and your position on a topic or issue.

3. AI is a tool that needs to be cited. If you use AI on an assignment, you need to acknowledge that in a paragraph describing how AI was used and listing the prompts used. Failure to provide this information is a violation of academic integrity. Your paragraph about AI use and prompts should be submitted a downloadable attachment in the 'Add Comment' feedback section of the assignment.

4. AI is not appropriate for all situations and contexts. Be thoughtful in how you use it. AI is not a replacement for knowing and understanding the material, but it can be helpful in getting you started, learning the information more deeply, and maybe even proof-reading and improving your papers. Be aware, however, that the texts and images produced by AI prompts are currently not protected by U.S. copyright law.

Rondrea Danielle Mathis: English, Bethune-Cookman University

ACADEMIC HONESTY

Academic honesty is essential to student success. Copying or paraphrasing another author’s work without proper documentation and acknowledgement amounts to plagiarism. It is also academically dishonest to purchase assignments/papers from other people.

Cheating, plagiarism, and all other forms of academic dishonesty are contrary to the policies of the university. By virtue of your student status at our university, you have proven that your own ideas and thoughts are more than sufficient to contribute to the body of scholarly work at this institution.

Therefore, the first instance of academic dishonesty will result in a failing grade for an assignment. Any subsequent cases or instances of academic dishonesty will result in a failing grade for the course, academic suspension and/ or dismissal from the university. If you are unsure about what constitutes cheating and/or plagiarism, please ask. Ignorance of the university’s policies will not absolve you from punishment.

Use of AI: Students may use AI to general base level ideas, clarify concepts, and conduct simple research, and students must properly cite their use of AI. Students may not use AI to generate their essays or assignments. I expect the assignments students submit to be their own work or writing, not an artificially intelligent large language model’s.

When indicating on a syllabus that chatbots and other AI tools are prohibited, instructors should aim for maximum clarity. They might also consider referencing the UNK Academic Integrity policy for further guidance.

General AI Statements

"In this course, integrity and original thought are paramount. Therefore, the use of generative artificial intelligence (AI) tools, including but not limited to AI-driven content creation services, is strictly prohibited in completing any course-related assignments, quizzes, examinations, projects, and papers. This policy is in place to ensure that all submitted work genuinely reflects individual effort and understanding of the course material.

Students are required to rely on their knowledge, course materials, and approved resources to complete their assignments. Usage of generative AI for drafting, revising, or generating content that is meant to be submitted as part of your coursework will be considered an infringement of academic integrity policies.

Violation of this policy will result in disciplinary actions which may include, but are not limited to, receiving a failing grade on the assignment, a failing grade for the course, and further disciplinary measures as outlined in the Academic Integrity Policy

If you are unsure whether a tool or resource falls under this prohibition, please consult with the instructor before proceeding with its use.

Alternative Statements

Students are not allowed to use generative AI tools or applications, such as ChatGPT and other AI writing or coding aids, for any assignments in this course.
Using generative AI tools, including ChatGPT and similar AI writing and coding assistants, to complete or assist in completing exams, term tests, assignments, or any other form of academic assessment may be deemed an academic offense in this course.
Students are forbidden from using any content derived from generative AI applications, including ChatGPT and other AI writing and coding assistants, whether by copying or paraphrasing, for their coursework in this course.
The employment of generative AI tools and applications, including ChatGPT and other similar AI writing and coding aids, is strictly forbidden in all coursework for this class unless specifically allowed by the instructor. Utilizing generative AI may be treated as using unauthorized aids and constitutes cheating.

In this course, students are encouraged to use artificial intelligence (AI) tools, such as ChatGPT, to reinforce class concepts and assist in the creation of study materials such as notes and outlines. This innovative approach aims to enhance understanding and engagement with the subject matter. However, to ensure transparency and academic integrity, the following guidelines must be observed:

Disclosure of AI Use: Students must inform the instructor about the use of any AI tool by submitting a detailed account of the prompts used. This disclosure should accompany the submission of the assignment where the AI tool was utilized.

Citation Requirements: All outputs generated from AI tools must be properly cited in APA format. This includes direct quotations, paraphrased material, and any substantive ideas derived from an AI-generated response. Failure to cite AI assistance will be considered a violation of the academic integrity policy.

Scope of Use: AI tools should be used to supplement student learning and assist in the creation of non-graded study materials only. AI-generated work should not replace the primary student effort in graded assignments and assessments unless explicitly authorized by the instructor.

Alternative Statements

Students are allowed to utilize generative AI tools, such as ChatGPT and other AI writing or coding aids, strictly for reinforcing course concepts and preparing study materials like notes and outlines. However, use of these tools for graded assignments or exams is not permitted.
Utilizing generative AI tools, including ChatGPT and similar AI writing and coding assistants, to aid in the preparation of notes or outlines must be clearly disclosed to the instructor, including detailed descriptions of the prompts used.
It is forbidden for students to submit any content directly derived from generative AI applications, including ChatGPT and other AI writing and coding assistants, as part of their graded coursework, unless explicitly authorized by the instructor.
The use of generative AI for creating non-graded educational materials is permitted; however, all generated content must be critically reviewed by the student to ensure accuracy and relevance to the course material. Misuse of AI in producing graded work will be treated as an academic integrity violation.

In this course, students are permitted to use artificial intelligence (AI) tools, including but not limited to AI content generators like ChatGPT, for the creation of course assignments and projects. This policy is designed to encourage exploration and integration of innovative digital tools in academic work. However, students must be aware of and adhere to the following responsibilities:

Accountability for Submitted Work: Students are fully responsible for all content submitted, including the accuracy and integrity of information, whether generated by AI tools or produced traditionally. It is the student's duty to verify the correctness of information and arguments produced by AI, and to ensure that the final submission adheres to academic standards.

Disclosure of AI Use: While using AI tools is allowed, transparency is crucial. Students should disclose the use of any AI tools in their submissions. This disclosure does not need to accompany each use but should be included in a general statement within the work submitted.

Critical Engagement: Students are encouraged to critically engage with and scrutinize the AI-generated content for any potential biases or inaccuracies. Relying solely on AI-generated content without personal oversight and critical assessment may negatively impact the quality and credibility of the work.

Adherence to Academic Standards: The use of AI must align with all existing academic integrity guidelines. Students must ensure that their use of AI tools does not contravene any guidelines regarding plagiarism and proper citation practices.

Alternative Statements

Students may freely use generative AI tools and applications, such as ChatGPT and other AI writing or coding aids, in the creation of their coursework. Students must ensure that all submitted work, whether generated by AI or not, is accurate and adheres to academic standards.
While employing generative AI tools in coursework is allowed, students must disclose the use of such tools in their submissions. This disclosure should be part of a general statement of AI use included in the submitted work.
Students must critically engage with all AI-generated content to identify and correct any inaccuracies or biases. Reliance on unchecked AI-generated content is strongly discouraged and could impact the grading of the work.
All uses of generative AI must comply with academic integrity standards. Failure to verify the accuracy or originality of AI-generated content could result in academic penalties for violating academic integrity policies.

Writing Effective Prompts

One of the keys for both you and your students to use generative AI effectively is to craft prompts that provide enough information for the tool to create a useful output. Such prompts should be precise and detailed, avoiding ambiguity. Whether it is you using AI to generate content or assignments for your class or your students using AI for an assignment, the user should clearly define the context and scope of the prompt, specifying any constraints or specific details that need to be included in the AI's response.

Tips for Writing AI Prompts

Be clear and concise
Harness chat threads
Talk to it like a colleague
Specificity matters
Be patient, revise prompts

Elements of an effective prompt

Task

Clearly state your request. Include in the prompt the topic, what you want it to do, and any specific parameters to the request.

Context

Provide the AI with any relevant background information related to the task.

Role

Direct the AI to act as a person in a specific role when generating your content

Audience

Detail who the audience is, what their knowledge level of the material is, and the tone you prefer.

Format

Content can be formatted in tables for Excel, ordered and unordered lists, paragraphs with headings, etc.

Sample Prompt to Write Learning Objectives

Role Context Task

Act as a university professor of American history. You are teaching an entry-level American history course. For the first module, students read oral narratives and creation myths from various Native American tribes. They also view artifacts and ruins. The goal is for students to have a strong understanding of Native life prior to colonization. Please write four learning objectives for this module.

Sample Prompt to Write a Blog Post

Role Context Task Format

Act as a professor of education. Your department has started a blog to discuss hot topics in education. Write a 300-word blog post discussing the top five ways educators can harness generative AI. Please include a title and organize the content with headers in an ordered list.

Sample Prompt to Brainstorm an Assignment

Role Context Task Audience

You are a professor at a university. You teach Public Relations 101 to online students. For most of your students this is their first exposure to Public/Media Relations. Module 6 covers the art of drafting an impactful press release and building robust medial relationships. Could you brainstorm 5 engaging and interactive learning activities you would assign to students to reinforce the module content?

Re-Prompting and Refining your prompts

When working with generative AI, refining your prompts can significantly enhance the quality and relevance of the output. Here are some tips for reprompting and refining your prompts:

Start Specific and Detailed

Initial Clarity: Begin with a clear and detailed prompt. The more specific you are, the better the AI can understand your request.
Example: Instead of asking, "Tell me about criminal justice," ask, "Explain the role of community policing in modern criminal justice systems."

Use Iterative Refinement

Iterative Approach: Start with a general prompt and refine it based on the AI's initial response.
Refinement Steps: If the first output is too broad, narrow the scope in your next prompt.
- Example:
  - First prompt: "Discuss community policing."
  - AI response: Provides a broad overview.
  - Refined prompt: "Discuss the benefits of community policing in reducing crime rates in urban areas."

Incorporate Examples and Context

Examples: Provide examples to guide the AI.
Context: Offer context to make your prompt more specific.
- Example: "Describe how community policing strategies have been implemented in New York City and their impact on local communities."

Ask Follow-Up Questions

Follow-Up: Use follow-up questions to delve deeper into specific aspects.
- Example: If the AI mentions several strategies, ask, "Can you elaborate on the community outreach programs mentioned?"

Use Constraints and Guidelines

Constraints: Set boundaries for the response length, format, or style.
- Example: "In 200 words, summarize the key components of community policing."

Clarify Ambiguities

Clarification: If the AI's response is unclear or off-topic, clarify your requirements.
- Example: "Your previous response was about community events. Can you focus on the collaboration between police and local businesses instead?"

Employ Step-by-Step Instructions

Step-by-Step: Break down complex prompts into smaller, manageable steps.
- Example:
  - Step 1: "List the primary goals of community policing."
  - Step 2: "Explain how each goal contributes to crime reduction."

Use Multiple Perspectives

Perspectives: Ask the AI to consider different viewpoints or perspectives.
- Example: "Compare the effectiveness of community policing in urban vs. rural settings."

Feedback Loop

Feedback: Provide feedback on the AI's output to refine future prompts.
- Example: "Your explanation was good, but can you provide more data or statistics to support your points?"

Refining your prompt

Original Prompt: You are a professor at a university. You teach Public Relations 101 to online students. For most of your students this is their first exposure to Public/Media Relations. Module 6 covers the art of drafting an impactful press release and building robust medial relationships. Could you brainstorm 5 engaging and interactive learning activities you would assign to students to reinforce the module content?

Output

Sure, here are five engaging and interactive learning activities for Module 6 of your Public Relations 101 class:

1. Press Release Simulation Exercise

Activity: Students will be divided into small groups and given a scenario (e.g., a product launch, a crisis situation, a corporate event). Each group will draft a press release based on the given scenario.

Objective: To practice writing clear, concise, and compelling press releases.

Engagement: Groups will present their press releases to the class, simulating a real-world press briefing.

2. Media Relationship Role-Playing

Activity: Students will pair up, with one acting as a PR professional and the other as a journalist. The PR professional will pitch a story to the journalist, who will ask questions and provide feedback.

Objective: To understand the dynamics of media relations and effective pitching.

Engagement: Pairs will switch roles and repeat the exercise with a different scenario, fostering a deeper understanding of both perspectives.

3. Press Release Critique Workshop

Activity: Students will bring in press releases from real companies (found online) and critique them in small groups. They will discuss what works well and what could be improved.

Objective: To develop critical thinking skills and understand common press release pitfalls.

Engagement: Each group will present one press release and their critique to the class, encouraging collaborative learning and discussion.

4. Virtual Guest Speaker Session

Activity: Invite a PR professional or journalist to speak to the class about their experiences with press releases and media relationships. Students will prepare questions in advance and participate in a Q&A session.

Objective: To gain insights from industry professionals and apply theoretical knowledge to real-world scenarios.

Engagement: The interactive Q&A allows students to engage directly with the speaker, deepening their understanding of the topic.

5. Press Release Distribution Strategy Project

Activity: Students will create a distribution plan for their press release, identifying key media outlets, crafting personalized pitches, and outlining a timeline.

Objective: To understand the importance of strategic planning in media relations.

Engagement: Students will present their distribution strategies to the class, receiving feedback from peers and the instructor. This activity encourages strategic thinking and practical application of course content.

These activities will help students apply the theoretical knowledge from the module in practical, engaging ways, reinforcing their learning and preparing them for real-world PR tasks.

Re-Prompt

While these outputs are creative and might lead to great learning opportunities for students, the PR class in question is asychronous, and online, so some of the ideas might not be feasible. Let's re-prompt the AI to point out this information and see if it can give us some more feasible options.

Re-Prompt: I love your ideas but unfortunately my class is online and asychronous, can you give me some ideas that will fit better with my class structure?

Certainly! Here are five engaging and interactive learning activities tailored for an online and asynchronous Public Relations 101 class:

1. Press Release Drafting and Peer Review

Activity: Students will draft a press release based on a given scenario (e.g., a product launch, a crisis situation, a corporate event). They will then submit their drafts to a discussion board where classmates can provide constructive feedback.

Objective: To practice writing clear, concise, and compelling press releases and improve through peer feedback.

Engagement: Students will review at least two peers' press releases, offering suggestions and comments, fostering a collaborative learning environment.

2. Virtual Media Relationship Role-Playing

Activity: Using a discussion board or a collaborative document, students will simulate a media pitch. One student will act as a PR professional pitching a story, and another will respond as a journalist with questions and feedback.

Objective: To understand the dynamics of media relations and effective pitching.

Engagement: Students will be assigned roles and scenarios, and they will post their interactions asynchronously, allowing for thoughtful responses.

3. Press Release Analysis Forum

Activity: Students will find a real press release online and post it to a forum along with their analysis of its strengths and weaknesses.

Objective: To develop critical thinking skills and understand common press release pitfalls.

Engagement: Students will comment on at least two analyses posted by their peers, discussing their thoughts and further insights.

4. Virtual Guest Speaker Video

Activity: Record an interview with a PR professional or journalist discussing their experiences with press releases and media relationships. Post the video for students to watch.

Objective: To gain insights from industry professionals and apply theoretical knowledge to real-world scenarios.

Engagement: Students will submit a reflection assignment or a discussion post with key takeaways from the video and questions they would ask the speaker, fostering engagement with the content.

5. Press Release Distribution Strategy Assignment

Activity: Students will create a distribution plan for their press release, identifying key media outlets, crafting personalized pitches, and outlining a timeline. They will submit their plans for instructor feedback.

Objective: To understand the importance of strategic planning in media relations.

Engagement: In a follow-up discussion board, students will share their distribution strategies and provide feedback on each other’s plans, promoting peer learning and interaction.

These activities are designed to fit within an asynchronous online class structure while maintaining engagement and interactivity.

View chat of these prompts

Continuing conversations to refine your results

One of the strengths of these tools is the ability to work with the tool within the ongoing context of a specific conversation. In this example, we have the same conversation as is linked above. We like the idea of a virtual media relationship roleplaying scenario laid out in the previous example but we are a little unsure of how this might actually be constructed. We can ask ChatGPT to write a same assignment description to help us picture how we might want to structure such an assignment in our class:

Prompt: I really like the Virtual Media Relationship Role-Playing idea. Can you write a sample assignment description? Remember this is an asynchronous class.

Virtual Media Relationship Role-Playing Assignment

Objective:

The purpose of this assignment is to help you understand the dynamics of media relations and practice effective pitching techniques in a simulated environment.

Instructions

Scenario Assignment:

You will be assigned a role (either PR professional or journalist) and a specific scenario. Scenarios may include a product launch, a crisis communication situation, a corporate event announcement, or a community outreach program.

PR Professional Role:

If you are assigned the PR professional role, your task is to craft a compelling pitch for the scenario. Your pitch should include:

A brief introduction of your company/organization.

Key details about the scenario/event.

The significance or newsworthiness of the story.

Any relevant quotes or supporting information.

A clear call to action for the journalist (e.g., request for an interview, attendance at an event).

Journalist Role:

If you are assigned the journalist role, your task is to respond to the PR professional’s pitch. Your response should include:

Questions seeking further information or clarification.

Feedback on the pitch (e.g., what caught your interest, what could be improved).

Your decision on whether to pursue the story, with reasoning.

Discussion Board Interaction:

PR professionals will post their pitches in the designated discussion board forum by [due date].

Journalists will respond to the pitches within 48 hours of the posting, providing questions and feedback.

Reflection and Analysis:

After completing the role-playing interaction, each student will write a brief reflection (300-500 words) on the experience. Reflect on what you learned about the pitching process, the interaction between PR professionals and journalists, and any insights gained from the feedback received or provided.

Submit your reflection to the assignment submission link by [due date].

Evaluation Criteria:

PR Pitch Quality (40%): Clarity, completeness, and persuasiveness of the pitch.

Journalist Response (40%): Thoughtfulness, relevance of questions, and constructive feedback.

Reflection and Analysis (20%): Depth of insight, personal learning outcomes, and connection to course material.

Submission Deadlines:

PR Pitch Posting: [due date]

Journalist Response: Within 48 hours of PR pitch posting

Reflection and Analysis: [due date]

Participation:

Active participation in both roles and thoughtful engagement in the discussion board are essential for a successful assignment. Ensure you adhere to the deadlines for each part of the assignment.

Assignment Examples

Below are examples of how UNK faculty are using Generative AI in their classes:

Christina Sogar-Department of Social Work

Group Theory Presentation utilizing AI in Human Behavior Analysis

Objective:

The objective of this assignment is to integrate artificial intelligence (AI) into the exploration and understanding of human behavior. Students will use AI to generate a unique case study and then apply a specific theory to analyze the behavior presented. This assignment aims to enhance critical thinking, theoretical application skills, and understanding of the interplay between technology and social work practice.

Assignment Details:

AI-Generated Case Study Creation:

Utilize an AI platform capable of generating realistic case scenarios based on specified parameters. You should define these parameters to ensure the case study focuses on relevant themes such as family dynamics, social issues, individual psychology, or community interactions. Include factors on all system levels: micro, mezzo & macro.
Parameters to consider might include age, socioeconomic status, cultural background, and specific challenges or stressors (e.g., loss of a job, substance abuse, migration).
Once generated, review the case study to ensure it is coherent, relevant, and appropriate for academic analysis. Modify if necessary to meet ethical and educational standards.
Must submit all prompts and responses with final presentation.

Theoretical Framework Selection:

Each group of 3 students will be assigned one of the following theoretical frameworks for analyzing the generated case study:

Humanistic Theory: Carl Rogers

Symbolic Interaction Theory: George Herbert Mead
Feminist Theories
Social Learning Theory: Albert Bandura

Provide a brief summary of the chosen theory, including its main concepts and how it applies to understanding human behavior. Include a critique of the theory assessing its strengths and weaknesses.

Case Study Analysis:

Apply your chosen theoretical framework to the AI-generated case study. Analyze the behaviors, relationships, and social structures presented in the scenario through the lens of your selected theory. Consider how the theory assists with your understanding of situational factors on the micro, mezzo and macro levels.
Discuss how the theory helps in understanding the complexities of the case, including the individual's behavior, social interactions, and potential interventions. How do the key elements of the theory contribute to assessing the interpersonal dynamics of the case?
Based on the key principles of the theory, recommend an intervention to assist the family members in enhancing their mental, emotional, social, and physical functioning, strengthening relationships, and dealing with any crisis situations.
Identify any limitations or challenges in applying your chosen theory to the case study.

Reflection:

Reflect on the process of using AI to generate a case study. Discuss the benefits and potential drawbacks of using artificial intelligence in social work practice, particularly in the context of understanding human behavior.
Consider the ethical implications, accuracy, and applicability of AI-generated scenarios in educational and professional settings.

Presentation:

Your assignment should be presented in a well-organized and engaging slide presentation. It should include an introduction, methodology (AI parameters and generation process), case study narrative, assessment and intervention analysis, and reflection sections.
Support your analysis and reflection with citations from relevant sources, including peer-reviewed articles, textbooks, and credible online resources as needed.
All group members must be actively involved in the case study generation, analysis and presentation. Peer reviews will be done following the in-class presentations.

Submission Details:

Due Date: Week 16 (5/6, 5/8)
Format: PowerPoint presentation including a reference slide

Assessment Criteria:

Clarity and relevance of the AI-generated case study.
Understanding and application of the selected theoretical framework.
Depth of analysis and integration of theory with the case study.
Quality and coherence of writing, including grammar and APA formatting.
Insightfulness of the reflection on the use of AI in social work practice.

Group Theory Presentation: Rubric

Category

Exemplary

AI-Generated Case Study

· 5 points

Case study is clear & presents detailed & relevant information.

Presented to class in no more than 2 minutes.

Includes description of methods used to generate case study narrative.

All prompts and responses submitted with final project.

Description of theoretical framework

· 15 points

Describes theory accurately & cohesively.

Provide brief background of development of theory.

Addresses strengths & weaknesses of theory.

Minimum of two scholarly sources required (not including text).

Utilization of theory

· 40 points

Illustrates critical application of theory to case study dynamics.

Description of case components is congruent with key theoretical concepts.

Considers role of multi-cultural and identity elements in application of theory.

Theory is used to inform assessment of case on micro, mezzo, & macro levels.

Recommendation of specific intervention plan related to theoretical analysis of case.

Consider limitations in applying theory to case study.

Reflection

· 15 points

Reflection should fully address the following questions:

· How was the experience using AI to generate a case?

· How many times did you revise the prompts? How was the process?

· What are potential ethical implications of incorporating AI into higher education as well as social work practice?

General presentation assessment criteria:

· Well-organized, clear & engaging;

· All required components included;

· Relevant scholarly sources correctly cited throughout presentation;

· Cohesive integration of theory to case assessment and intervention;

· Active involvement of all group members.

Daniel Chaffin: Department of Management

Complete the following Porter's Five Forces analysis table for the INDUSTRY your company competes in using Bing Chat or some other AI. Attempt using creative mode if possible. Be sure to include footnotes to support your assertions (at least 5 footnotes - be sure to check the footnotes that they support your assertions and include them).

To complete this assignment you need to break it down step by step.

Start with figuring out what industry your company is in. You can Google this. For example, what industry is "company" in? You could also ask, "what is the NAICS code for the _____ industry?"
Ask something like, "Using Porter's five forces as an analytical tool, what is the bargaining power of the threat of new entrants for the ____ industry? Justify your answer with a compelling rationale and evidence with footnotes"
Maybe next ask something like, "your rationale is weak. Further develop your rationale utilizing theory and evidence and give 4 compelling reasons"
Next ask something like this, "Assume you are a company owner and strategist for ____ firm in the ____ industry, what are four strategies you would you consider to lower the threat of new entrants? Support these strategies with evidence and footnotes"
Next say something like this, "those strategies are pretty mundane and generic. Develop stronger, distinctive and creative strategies. Include links and sources to support your ideas"

Now repeat it for all of the cells.

Force	Level (high, medium or low)	Rationale	Company strategies that increase the bargaining power of this force	Company strategies that lower the bargaining power of this force
New entrants
Buyer
Supplier
Substitutes
Rivalry

Submit a file with your completed table and references. Also include answers to the following questions:

Where did the AI mess up in your analysis?
What did you do to correct it?
What did the AI get correct in its analysis that you didn't think of?
What in your prompts helped you get better answers to complete the table?

Here is a video to illustrate:

;

Amanda Sladek-Department of English and Modern Languages

Writing Project 2: AI-“Enhanced” Essay

Task

“Cheat” on this research essay assignment by using generative AI (ChatGPT, Google Gemini, Jenni, etc.) to assist. This means you can copy and paste AI-generated text directly into your essay. But keep careful track of what AI-generated text you use, the prompts you use, and the text generated, since you will also turn in a “revealed” copy with the AI-generated text highlighted, along with a cover letter reflecting on your experience.

Purpose

This unit and project will develop your critical awareness of Artificial Intelligence, especially its capacity to assist with writing, make writing worse, and make the whole process really ethically complicated. You will also build your research skills, including using library resources, evaluating sources, and reading academic articles. Finally, we will work on developing thesis-driven academic arguments, much like the kind you read in academic journals. Fulfills all LOPER 2 outcomes.

Suggested Steps

Start with a research question. You may begin by searching some publications in your major or area of interest to get a sense of the persistent questions, issues, or debates. You may already have an idea of a topic you’re interested in. Either way, this question should be something that you’re genuinely interested in and that you don’t already know the answer to/have a set opinion about. This will make it easier to keep an open mind as you approach your research!

Research your question. Your final paper must cite 5 sources, and at least 3 of them must be peer reviewed. If you use AI to help you research, be sure to save any prompts you use and responses generated so you can draw on them in your cover letter.

Write your AI-assisted draft. This will be a thesis-driven academic argument paper, similar to what you would see in a scholarly journal. You are allowed to use AI tools however you want in this paper, including copy/pasting text it generates for you, as long as the final product contains at least 40% original writing.1 Your paper should be around 2000 words, not including citations (but including AI-generated text). As you use AI, be sure to:

keep track of what text is AI-generated (for your AI revealed version) and

save any prompts you use and responses generated so you can draw on them in your cover letter.

Write your cover letter. Your cover letter should be approximately 1000 words, addressed to me, and answer several of the following questions (starred questions are required):

Describe how you used AI in the creation of this paper (research, drafting, editing, etc.). Did it help you in any ways that aren’t clear just from looking at the highlighted text in the AI revealed copy? Include some of the specific prompts you used and the responses it gave you.*

Looking at your paper, what differences do you notice between the AI-generated text and your own writing?*

Do you think incorporating AI made your writing better or worse (or both)? Why?*

Did the use of AI help your writing process? Why or why not?

Do you think the use of AI contributed to or hindered your learning (or both)? Why?*

Do you feel like you “cheated” on this assignment? Why or why not?*

If you could write an AI policy to be included in UNK course syllabi, what would it look like?

What did you learn about AI in this unit and assignment?*

Audience

When writing your papers, your audience should be academic readers in your chosen field (picture readers of a journal published in your discipline, except please don’t use AI when writing for actual journals). The cover letter should be addressed to me.

You Will Turn In

Project 2 final draft (submit via Canvas)

AI concealed version, with no indication of what was written with AI assistance

AI revealed version, with AI-generated text highlighted

Cover letter

Project 2 reflection questions (fill out on Canvas)

Peer review (will be done online via Canvas)

Important Dates

March 11-15: Spring Break
Tuesday, March 19: Class will meet via Zoom
Thursday, March 28: Project 2 peer review, rough draft due
Friday, March 29: Final draft and Project 2 reflection due at 11:59 pm

This project was adapted from:

Fyfe, Paul. “How to Cheat on Your Final Paper: Assigning AI for Student Writing.” AI & Society, vol. 38, no. 4, 2023, pp. 1395–405, https://doi.org/10.1007/s00146-022-01397-z.

Bias in Generative AI

Bias in AI Overview

Generative AI models are susceptible to biases stemming from their training data and design.

When it comes to generative AI, it's crucial to acknowledge how these unconscious associations can influence models, leading to biased outputs. By identifying bias sources and crafting inclusive prompts, we can develop AI systems that are more fair and just.

Stages Where Bias Can Occur

Bias can occur at various stages of the AI pipeline:

Data Collection: Bias often originates here. If the training data is not diverse or representative, the AI algorithm might produce biased outputs.
Data Labeling: Bias can be introduced if annotators have different interpretations of the same label.
Model Training: This critical phase can result in biased outputs if the training data is unbalanced or the model architecture cannot handle diverse inputs.
Deployment: Bias can also be introduced if the system is not tested with diverse inputs or monitored for bias after deployment.

Implicit and Explicit Bias

It's essential to differentiate between explicit and implicit biases. Explicit biases are conscious prejudices or beliefs toward a specific group. In contrast, implicit biases are unconscious attitudes or stereotypes that influence our understanding, actions, and decisions without our awareness. These biases stem from in-group preferences and broader structural inequities, often absorbed from dominant social messages.

Implicit biases are particularly harmful because they are deeply ingrained and can influence our behavior even when we consciously reject them. To reduce bias and promote inclusivity and respect, we should be mindful of our language and actively seek diverse perspectives.

Types of Bias in AI

Biases can have severe repercussions, especially when they contribute to social injustice or discrimination. Biased data can exacerbate existing prejudices, leading to systemic inequalities. Therefore, it is crucial to detect and rectify biases in data and models to ensure fairness and impartiality in data-driven decision-making processes.

Selection Bias: Occurs when training data is not representative of reality. For example, if a dataset includes only male employees, the model cannot accurately predict female employees' performance.
Confirmation Bias: Happens when an AI system relies too much on pre-existing beliefs or trends, reinforcing existing biases and failing to identify new patterns.
Measurement Bias: Arises when collected data systematically differs from actual variables of interest. For instance, if a model predicts students' success in an online course using data only from students who completed the course, it may not accurately predict the performance of those who dropped out.
Stereotyping Bias: Occurs when an AI system reinforces harmful stereotypes, such as a facial recognition system being less accurate in identifying people of color or a language translation system associating certain languages with specific genders or stereotypes.
Out-group Homogeneity Bias: Happens when an AI system struggles to distinguish between individuals not part of the majority group in the training data, leading to misclassification or inaccuracy when dealing with minority groups.

Implications for Higher Education

- Content Generation: Biased AI-generated content can perpetuate stereotypes and misinformation in educational materials, impacting students' understanding and perspectives.
- Assessment and Feedback: If used for grading or providing feedback, biased AI models may unfairly evaluate students' work, disadvantaging certain groups and affecting their academic performance.
- AI Detection Tools: AI detection tools such as Turnitin's AI technology can produce false positives, incorrectly identifying students' original work as generated by AI, which can unfairly penalize students and damage their academic integrity. This issue can be particularly biased against ESL students, whose writing might deviate from native linguistic norms, leading to a higher likelihood of being misclassified as AI-generated, thereby exacerbating stress and undermining their academic progress.
- Research Assistance: AI tools that exhibit bias can influence the direction and quality of academic research, leading to skewed findings and perpetuating existing biases in scholarly work.

Additional Resources

Does AI have a Bias Problem?

What do we do about Bias in AI

Alan Turing Institute-Understanding AI Ethics and Safety

Copyright Considerations

Issues related to generative AI and copyright can be divided into two distinct categories:

These questions center around who owns the inputs to the model, whether this is considered fair use, and who owns the outputs of generative AI.

Does unauthorized use of copyrighted material in model training violate the copyright act?

Maybe..legal challenges are still on going including suits against chip giant nvidia and the New York Times vs Open AI.

Does the use of copyrighted materials in model training constitute Fair Use?

AI companies may argue that their training processes constitute fair use and are therefore noninfringing. Whether copying constitutes fair use depends on four statutory factors under 17 U.S.C. § 107:

The purpose and character of the use, including whether it is commercial or for nonprofit educational purposes.
The nature of the copyrighted work.
The amount and substantiality of the portion used in relation to the copyrighted work as a whole.
The effect of the use on the potential market for or value of the copyrighted work.

Some stakeholders argue that using copyrighted works to train AI programs should be considered fair use under these factors. For the first factor, OpenAI claims its purpose is "transformative" rather than "expressive" because the training process creates "a useful generative AI system." OpenAI also argues that the third factor supports fair use since the copies are not made public but are used only for training the program. OpenAI cites The Authors Guild, Inc. v. Google, Inc., where the U.S. Court of Appeals for the Second Circuit held that Google's copying of entire books to create a searchable database that displayed excerpts constituted fair use.

Regarding the fourth fair use factor, some generative AI applications have raised concerns that training AI programs on copyrighted works allows them to generate similar works that compete with the originals. For example, an AI-generated song called “Heart on My Sleeve,” made to sound like the artists Drake and The Weeknd, was heard millions of times on streaming services. Universal Music Group, which has deals with both artists, argues that AI companies violate copyright by using these artists’ songs in training data. OpenAI states that its visual art program DALL-E 3 "is designed to decline requests that ask for an image in the style of a living artist."

Are the works I or my students create using generative AI protected by copyright?

No, probably not. According to the guidance given by the US Copyright Office in March 2023, Human authorship is necessary for a material to be copyrighted:

In the Office's view, it is well-established that copyright can protect only material that is the product of human creativity. Most fundamentally, the term “author,” which is used in both the Constitution and the Copyright Act, excludes non-humans. The Office's registration policies and regulations reflect statutory and judicial guidance on this issue.

Additional Resources

Artificial Intelligence and CC licenses