Artificial Intelligence subjectId: 690674

Welcome to the Artificial Intelligence Research Guide! Artificial Intelligence (AI) is a field dedicated to developing computers and robots capable of imitating the human mind. The term “AI” was coined by John McCarthy during a 1956 summer workshop at Dartmouth University. However, many technical concepts that define AI predate this period. Alan Turing’s theoretical ideas significantly influenced future AI research.

Since 1956, AI has evolved, and more recently, there has been a surge in its use, particularly with the creation of ChatGPT in spring 2023. While the exact future of AI remains uncertain, it is undeniable that AI will most likely continue to impact every aspect of our lives. This guide provides information on starting your AI research, with available resources at the University of Miami Libraries and beyond. 

This guide is not exhaustive, and the field of AI is dynamic. If you have any comments or questions please email me.

*Created using Microsoft Copilot

Appendix I: A short history of AI | One hundred year study on artificial intelligence (AI100). (n.d.). Retrieved July 26, 2024, from https://ai100.stanford.edu/2016-report/appendix-i-short-history-ai

Artificial Intelligence (AI) vs. Machine Learning. (n.d.). CU-CAI. Retrieved July 26, 2024, from https://ai.engineering.columbia.edu/ai-vs-machine-learning/

Computer Science
Robotics

Algorithm: a sequence of steps to be followed in calculations or other problem-solving operations.

Artificial Intelligence (AI): technology that allows computers and machines that can think and make decisions autonomously without human intervention. 

ChatGPT: natural language processing chatbot powered by generative AI that enables human-like conversations to complete tasks.

Deep Learning: subset of machine learning that utilizes multi-layered neural networks to simulate the human brain's ability to perform complex decision-making.

Generative AI: a category of artificial intelligence that can create new data that mimic a given set of training data and can produce outputs such as images, music, texts, and videos.

GPT (Generative Pre-trained Transformers): an artificial intelligence (AI) tool that can produce clear, coherent, human-like text and computer code in response to a question or prompt submitted by a user.

Hallucination: a large language model (LLM) perceives patterns or objects that are nonexistent or imperceptible to human observers, creating outputs that are nonsensical or altogether inaccurate.

Large Language Model (LLM): a category of foundation models trained on immense amounts of data making them capable of understanding and generating natural language and other types of content to perform a wide range of tasks.

Machine Learning (ML): a subfield of artificial intelligence that uses algorithms trained on data sets to create models that enable machines to perform tasks that would otherwise only be possible for humans, such as categorizing images, analyzing data, or predicting price fluctuations.

Natural Language Processing (NLP): a form of artificial intelligence (AI) that allows computers to understand human language, whether it be written, spoken, or even scribbled. 

Prompt: a question, command, or statement that you input into an AI model to initiate a response or action.

Prompt Engineering: the process of iterating a generative AI prompt to improve its accuracy and effectiveness.

Reinforcement Learning: A type of dynamic programming that trains algorithms using a system of reward and punishment. In reinforcement learning, an agent takes actions in a specific environment to achieve a predetermined goal, learning from the outcomes of its actions to maximize cumulative rewards.

Self-Supervised Machine Learning: A type of machine learning where models train themselves on unlabeled data. This approach, also known as predictive or pretext learning, allows algorithms to learn one part of the input from another part, automatically generating labels and transforming unsupervised problems into supervised ones.

Semi-Supervised Learning: A type of machine learning that combines elements of both supervised and unsupervised learning. It involves training algorithms on a small labeled dataset and a larger unlabeled dataset. The labeled data helps guide the learning process for the unlabeled data, making it useful for tasks where labeling data is expensive or time-consuming.

Supervised Machine Learning: A type of machine learning where the model is trained on a labeled dataset. This approach is commonly used for tasks such as risk assessment, image recognition, predictive analytics, and fraud detection.

Unsupervised Machine Learning: A type of machine learning where algorithms draw inferences from unlabeled datasets. This approach is used for exploratory data analysis, pattern recognition, and predictive modeling.

*Created using Microsoft Copilot

Artificial intelligence (AI) algorithms: A complete overview | Tableau. (n.d.). Retrieved June 3, 2024, from https://www.tableau.com/data-insights/ai/algorithms

Artificial intelligence tools at the U | University of miami information technology. (n.d.). Retrieved June 3, 2024, from https://it.miami.edu/about-umit/resources/ai-tools/index.html

ChatGPT | Definition & facts | Britannica. (n.d.). Retrieved June 3, 2024, from https://www.britannica.com/technology/ChatGPT

Facebook, V. S. S. at, Instagram, V. S. S. at, LinkedIn, V. S. S. at, X, V. S. S. at, Flickr, V. S. S. at, Youtube, V. S. S. at, TikTok, V. S. S. at, University, C. S., Compliance, Safety, C., Statement, A., Statement, P., IX, T., & Comments. (n.d.). Glossary of AI terms. Retrieved June 3, 2024, from https://www.csus.edu/information-resources-technology/ai/glossary-of-ai-terms.html

Introducing ChatGPT. (n.d.). Retrieved June 3, 2024, from https://openai.com/index/chatgpt/?ref=blog-what-is-chat-gpt-understanding¶ms=ref-blog-what-is-chat-gpt-understanding

TLS, U. L. (n.d.). LibGuides: Artificial intelligence (AI): Introduction. Retrieved June 3, 2024, from https://guides.lib.utexas.edu/c.php?g=1363366&p=10070740

What are AI hallucinations? | IBM. (2023, September 1). https://www.ibm.com/topics/ai-hallucinations

What are AI prompts? (& How to make them better) | Copy.ai. (n.d.). Retrieved June 3, 2024, from https://www.copy.ai/blog/what-are-ai-prompts

What are large language models (LLMs)? | IBM. (2023, November 2). https://www.ibm.com/topics/large-language-models

What is ChatGPT? (n.d.). Retrieved June 3, 2024, from https://www.brandeis.edu/teaching/chatgpt-ai/index.html

What is ChatGPT and why does it matter? Here’s what you need to know. (n.d.). ZDNET. Retrieved June 3, 2024, from https://www.zdnet.com/article/what-is-chatgpt-and-why-does-it-matter-heres-everything-you-need-to-know/

What is generative AI? | IBM. (2024, March 22). https://www.ibm.com/topics/generative-ai

What is machine learning? Definition, types, and examples. (2024, March 27). Coursera. https://www.coursera.org/articles/what-is-machine-learning

What is natural language processing? Definition and examples. (2024, March 19). Coursera. https://www.coursera.org/articles/natural-language-processing

Use the below databases and eJournals to search for articles regarding Artificial Intelligence.

Tools listed below are either free, freemium, or accessible via subscription for University of Miami faculty and staff. Many of the tools listed require you to create an account. Available tools on this guide can be used to enhance your educational experience, foster collaborative projects, and drive groundbreaking discoveries. Step into the world of AI and unlock the full potential of your intellectual pursuits.

Disclaimer: The tools included in this guide are intended to provide an overview of various AI tools that may be beneficial for university students, faculty, and staff. The inclusion of any specific tool does not constitute an endorsement. Users are encouraged to independently verify the suitability, reliability, and appropriateness of each tool for their specific needs.

*Created using Microsoft Copilot

Heuristica

​AI-powered mind maps and concept maps for visual learning, thinking and research.

My Lens

MyLens AI visually summarizes the key insights you need.

ChatGPT

A large language model (LLM) chatbot created by OpenAI.

Claude

A large language model (LLM) chatbot built by Anthropic.

Copilot

Generative artificial intelligence chatbot developed by Microsoft.

Gemini

Generative artificial intelligence chatbot started by Google.

Dall-E 2

AI system that can create realistic images and art from a description in natural language.

Firefly

Create AI high-quality images with better composition, photorealistic details, and improved mood and lighting.

SlidesAI

Create Presentation Slides with AI in seconds.

Slidesgo

With a few clicks, you’ll have wonderful slideshows that suit your own needs.

SlideSpeak

Create presentations with AI, summarize PowerPoint or Word documents and much more.

ChatPDF

Chat with PDF in a conversational manner.

Wordtune

The ultimate AI writing tool that rewrites, rephrases, and rewords your writing.

ResearchRabbit

A scholarly publication discovery tool supported by AI.

Elicit

Search, summarize, extract data from, and chat with over 125 million papers.

Perplexity.ai

Uses natural language processing and machine learning to generate responses from a variety of reputable sources, citing links within the text.

SCISPACE

Save parts of a PDF or Copilot response as notes.

Prompt Engineering is the process of iterating a generative AI prompt to improve its accuracy and effectiveness.

Think of prompt engineering like giving really good directions. It’s like telling your friend exactly how to get to a party, including where to turn and what landmarks to look out for. In the same way, you give an AI clear instructions on what you want it to do or say. You have to be specific, like saying “take a left at the big palm tree” instead of just “go that way.” This helps the AI understand and give you exactly the kind of answer or help you’re looking for.

*Created using Microsoft Copilot

​​​​​

The CLEAR framework is a set of principles designed to improve the interaction with AI through prompt engineering. It stands for Concise, Logical, Explicit, Adaptive, and Reflective.

1. Concise
   • Be Direct: Use clear and straightforward language.
   • Eliminate Fluff: Remove unnecessary words that don’t add value to the prompt.
2. Logical
   • Order Matters: Structure your prompts in a logical sequence.
   • Step-by-Step: Break down complex tasks into simpler steps.
3. Explicit
   • Be Specific: Clearly state what you want the AI to do.
   • Avoid Assumptions: Don’t expect the AI to “read between the lines.”
4. Adaptive
   • Flexibility: Be ready to adjust your prompts based on the AI’s responses.
   • Iterative Approach: Refine your prompts through trial and error.
5. Reflective
   • Feedback Loop: Use the AI’s responses to reflect on the effectiveness of your prompts.
   • Continuous Improvement: Aim to improve the clarity and efficiency of your prompts over time.

*Created using Microsoft Copilot

"is representing the words or ideas of someone else as your own. Examples include, but are not limited to, failing to properly cite direct quotes and failing to give credit for someone else's ideas."

Honor Code

1. AI should help you think, not think for you. AI tools may be used to help generate ideas, frame problems, and perform research. It can be a starting point for your own thought process, analysis, and discovery. Do not use them to do your work for you, e.g., do not enter an assignment question into ChatGPT and copy and paste the response as your answer.
2. The use of AI must be open and documented. The use of any AI in the creation of your work must be declared in your submission and explained. Your faculty can provide guidance as to the format and contents of the disclosure. The undeclared use of any AI (including text, images, program code, musical notation, etc) in any work may be considered as plagiarism.
3. Engage with AI Responsibly and Ethically. Engage with AI technologies responsibly, critically evaluating AI-generated outputs and considering potential biases, limitations, and ethical implications in your analysis and discussions. Ensure that the data used for AI applications are obtained and shared responsibly. Never pass off as your own work generated by AI.
4. You are 100% responsible for your final product. You are the user; if the AI tool makes a mistake, and you use it, then it’s your mistake. If you don’t know whether a statement about any item in the output is true, then it is your responsibility to research it. If you cannot verify it as factual, you should delete it. You hold full responsibility for AI-generated content. Ideas must be attributed, and sources must be verified.
5. These principles are in effect unless the instructor gives you specific guidelines for an assignment or exam. It is your responsibility to ensure you are following the correct guidelines. Not following them will result in a breach of the Academic Integrity Policy.
6. Data that are confidential or personal should not be entered into generative AI tools. Putting confidential or personal data into these tools exposes you and others to the loss of important information. Therefore, do not do so. See point 3 above.
7. The rules and practices on the use of AI may vary from class to class, discipline to discipline. Do not assume that what is acceptable in a Computer Science class will be acceptable in a Philosophy class. It is the student’s responsibility to stay informed as to the instructor’s expectations. When in doubt, ask.

Teaching and Learning with AI from PETAL

In academia, researchers must remain vigilant when using AI, considering both its potential benefits and downsides. By addressing these challenges, you can harness AI’s power while minimizing its risks. This tab is dedicated to describing the disadvantages associated with artificial intelligence.

While AI has amazing capabilities, it’s essential to recognize its negative aspects such as:

• Bias
• Environmental Impact
• Hallucinations
• Labor
• Privacy

*Created using Microsoft Copilot

Users should be aware of the different biases in AI:

• Selection Bias
  • When the participants selected for a study are not representative of the target population, often due to non-random sampling or retention methods. This can lead to skewed results and inaccurate conclusions.
• Confirmation Bias
  • Confirmation bias is the tendency to seek out, interpret, and favor information that confirms one’s pre-existing beliefs or hypotheses. This bias can lead to flawed decision-making and reinforce existing misconceptions.
• Measurement/Information Bias
  • Occurs when key study variables are inaccurately measured or classified, leading to systematic errors in the study’s results. This bias can significantly distort the association between variables and affect the study’s validity.
• Stereotyping Bias
  • Involves forming generalized beliefs or assumptions about individuals based solely on their membership in a particular group, often ignoring their unique characteristics. This cognitive bias can lead to unfair treatment and reinforce existing prejudices.
• Out-group homogeneity bias
  • Tendency to perceive members of an out-group as more similar to each other than members of one’s in-group, often overlooking individual differences. This bias can lead to stereotyping and misjudgments about out-group members.

*Created using Microsoft Copilot

Artificial Intelligence (AI) demands significant energy, necessitating numerous servers in data centers, which in turn require substantial power. This leads to various environmental impacts, including:

• High Water Consumption: Data centers consume large amounts of water to cool servers.
• Air Pollution: Increased energy use can lead to higher emissions of air pollutants.
• High Energy Use: This often results in greater reliance on fossil fuels, contributing to greenhouse gas emissions.
• Inefficient Resource Use: The operation of data centers can lead to inefficient use of natural resources.
• E-Waste: The rapid advancement of technology can increase electronic waste

*Created using Microsoft Copilot

AI hallucinations refer to instances where artificial intelligence systems generate outputs that are false or inaccurate, often due to limitations in their training data or inherent design flaws. These hallucinations can undermine the reliability of AI tools. To combat AI hallucinations:

• Critically Evaluate Outputs: Always assess the information provided by AI critically.
• Diversify Sources: Use a variety of sources to cross-check and validate AI outputs.
• Stay Vigilant: Remain alert and cautious when interpreting results from any AI tool.

*Created using Microsoft Copilot

Note: While the term "hallucination" is commonly used in the AI industry, there are concerns that the phrase can possibly be undermining the "efforts to reduce stigma in psychiatry and mental health."

Artificial Intelligence (AI) is reshaping the labor market in profound ways, presenting both opportunities and challenges. On one hand, AI can enhance productivity and create new job categories, particularly in technology-driven sectors. However, it also poses the risk of job displacement, as AI systems can perform tasks traditionally done by humans, potentially leading to lower labor demand, wage reductions, and less hiring. This dual impact is more pronounced in routine tasks susceptible to automation, while jobs requiring complex problem-solving and creative skills may see growth. The integration of AI into the workplace can exacerbate existing inequalities, as those with higher education and specialized skills may be better positioned to benefit from AI, leaving less-skilled workers vulnerable.

*Created using Microsoft Copilot

The integration of Artificial Intelligence (AI) into various aspects of daily life has raised significant privacy concerns. One of the primary issues is the scale of data collection; AI systems require vast amounts of data to learn and improve, leading to an unprecedented level of personal data being harvested and analyzed. This often occurs without explicit user consent or awareness, resulting in a loss of control over personal information. Additionally, there’s the risk of reidentification and deanonymization, where AI can potentially track and identify individuals across multiple platforms, breaching the users' expectation of anonymity.

*Created using Microsoft Copilot

HIPAA

"the generative AI tool ChatGPT is not HIPAA compliant, and the University does not permit its use with any patient data. OpenAI, the company behind ChatGPT, also warns against inputting confidential data into the platform, which could constitute a public disclosure and lead to a loss in our ability to protect University information, including intellectual property." -Artificial Intelligence Tools at the U

"Copilot is not HIPAA compliant yet." -Artificial Intelligence Tools at the U

AI Detectors

"The use of systems that claim to detect AI-generated text (e.g. GPTZero, Copyleaks) are not recommended. Submitting students’ coursework to these systems may constitute a FERPA violation since students’ work is considered an educational record." -PETAL

Citing sources is an essential activity in academia and provides proper credit to authors. It allows your reader to find sources you used and also helps combat plagiarism. Citing AI is just as important as citing any other source.

Note: A majority of the text below is copied from APA , APA Style, The Chicago Manual of Style Online, and the MLA Style Center.

• When a generative AI model is used in the drafting of a manuscript for an APA publication, the use of AI must be disclosed in the methods section and cited.
• AI cannot be named as an author on an APA scholarly publication.
• When AI is cited in an APA scholarly publication, the author must employ the software citation template, which includes specifying in the methods section how, when, and to what extent AI was used. Authors in APA publications are required to upload the full output of the AI as supplemental material

APA  & APA Style

References: OpenAI. (2023). ChatGPT (Mar 14 version) [Large language model]. https://chat.openai.com/chat

In-text example: (OpenAI, 2024)

APA  & APA Style

• Credit AI-generated text when you reproduce its words within your own work.
• "You do need to credit ChatGPT and similar tools whenever you use the text that they generate in your own work. But for most types of writing, you can simply acknowledge the AI tool in your text (e.g., “The following recipe for pizza dough was generated by ChatGPT”).

The Chicago Manual of Style Online

Text generated by ChatGPT, OpenAI, March 7, 2023, https://chat.openai.com/chat.

The Chicago Manual of Style Online

• Cite a generative AI tool whenever you paraphrase, quote, or incorporate into your own work any content (whether text, image, data, or other) that was created by it. 
• Acknowledge all functional uses of the tool (like editing your prose or translating words) in a note, your text, or another suitable location. 
• Take care to evaluate the secondary sources it cites.
• It is not recommended to treat the AI tool as an author.

MLA Style Center

• Paraphrased in Your Prose: While the green light in The Great Gatsby might be said to chiefly symbolize four main things: optimism, the unattainability of the American dream, greed, and covetousness (“Describe the symbolism”), arguably the most important—the one that ties all four themes together—is greed.

• Works-Cited-List Entry: “Describe the symbolism of the green light in the book The Great Gatsby by F. Scott Fitzgerald” prompt. ChatGPT, 13 Feb. version, OpenAI, 8 Mar. 2023, chat.openai.com/chat.

• Creative Visual Works: Fig. 1. “Pointillist painting of a sheep in a sunny field of blue flowers” prompt, DALL-E, version 2, OpenAI, 8 Mar. 2023, labs.openai.com/.

• Quoting Creative Textual Works: “The Sunflower” villanelle about a sunflower. ChatGPT, 13 Feb. version, OpenAI, 8 Mar. 2023, chat.openai.com/chat.

MLA Style Center

Artificial Intelligence (AI) is revolutionizing the landscape of education, offering innovative tools and methods to enhance both teaching and learning experiences. This research guide aims to explore the multifaceted applications of AI in education. It provides educators with ideas on how AI can be effectively integrated into educational practices to foster a more engaging, efficient, and inclusive learning environment.

*Created using Microsoft Copilot

Artificial Intelligence (AI) tools for teaching

Copilot

Generative artificial intelligence chatbot developed by Microsoft.

Gradesope

Helps you seamlessly administer and grade all of your assessments.

Generative AI Product Tracker

Lists generative AI products that are either marketed specifically towards postsecondary faculty or students or appear to be actively in use by postsecondary faculty or students for teaching, learning, or research activities.

Developing an AI Syllabus Statement & Driving Class AI Discussion (NCSU)

Developing Syllabus Statements for AI (Tufts University)

Possible AI Syllabus Statements (Brandeis University)

Is it possible for students to work on your assignments using genAI tools?

Teach with Generative AI (Harvard)

Artificial Intelligence Teaching Guide (Stanford)

"The use of systems that claim to detect AI-generated text (e.g. GPTZero, Copyleaks) are not recommended. Submitting students’ coursework to these systems may constitute a FERPA violation since students’ work is considered an educational record."

PETAL

10 Ways Artificial Intelligence Is Transforming Instructional Design

Artificial Intelligence In Education: Teachers’ Opinions On AI In The Classroom

How AI Is Reshaping Higher Education