This guide will help you understand how AI works, its uses as a research tool, and how it fits into your research strategies. There are a lot of other applications of AI - for example as a writing assistant, or to create images. But this guide will only discuss AI as it works in the research process.
Remember, each of your instructors will choose how or when you can use AI in the work for your assignments. This guide can't answer the question of whether or not it is ok to use AI in a certain assignment. Only your instructor can guide you on that question. You should also refer to the Academic Integrity clause of the handbook for guidance on understanding your responsibility to avoid plagiarism.
AI (Artificial Intelligence), ChatBots, and AI assistants are a new and rapidly changing technology. Partially this is because AI tools have exceeded expectations in how well they can understand and process natural language and execute commands without the user needing to understand code or a set of tools. AI works by taking in a huge amount of data (words, images, lists, questions and answers, articles) from different parts of the internet and learning the patterns in the data. This process is referred to as "Machine Learning" . Machine learning algorithms detect patterns in large data sets and learn to make predictions by processing data, rather than by receiving explicit programming instructions (McKinsey & Company, 2023). Once they learn to match patterns they can begin to understand how to create the desired output to human questions and requests. Most AI tools are also the result of a huge amount of human work to intervene to teach and correct the AI programs and guide them towards correct pattern matching (Simonite, 2023).
The code and algorithms that create AI programs can be very complex, and most are secret because they are the property of private owners. But we do know that AI tools need to train on and use large amounts of data from the internet to work. However, it is important to know that currently no AI tool can access and use all of the data from the internet. Each AI tool pulls from different subsets of data that its creators made available to it. The quality and source of the data used is crucial to understanding and evaluating the quality of the results generated by AI tools. Datasets that do not have access to scholarly materials will not be accurate or suitable for academic research. Datasets that contain biased sources will replicate that bias in the AI generated results (Simonite, 2023).
This guide is meant to help you understand how AI generates results and how those results can be evaluated for your research needs. We will also give you some guidance on different ways to use AI in your research process and where you can find more information on learning how to effectively prompt or create questions for AI.
Artificial Intelligence - Artificial intelligence is a machine’s ability to perform the cognitive functions we usually associate with human minds. (McKinsey)
Human in the Loop—"Human-in-the-loop is a branch of AI that brings together AI and human intelligence to create machine learning models. It’s when humans are involved with setting up the systems, tuning and testing the model so the decision-making improves, and then actioning the decisions it suggests. “(Faculty.ai)
Deep Learning – “[Deep Learning] involves passing data through webs of math loosely inspired by the working of brain cells that are known as artificial neural networks. As a network processes training data, connections between the parts of the network adjust, building up an ability to interpret future data.”(Wired)
Generative AI – “Generative AI is a catch-all term for AI that can cobble together bits and pieces from the digital world to make something new—well, new-ish—such as art, illustrations, images, complete and functional code, and tranches of text that pass not only the Turing test, but MBA exams. Tools such as OpenAI’s Chat-GPT text generator and Stable Diffusion’s text-to-image maker manage this by sucking up unbelievable amounts of data, analyzing the patterns using neural networks, and regurgitating it in sensible ways.”(Wired)
Hallucinations—"AI hallucinations are incorrect or misleading results that AI models generate. These errors can be caused by a variety of factors, including insufficient training data, incorrect assumptions made by the model, or biases in the data used to train the model.”(Google Cloud)
Large Language Models—"A language model is a machine learning model that aims to predict and generate plausible language. Autocomplete is a language model, for example. These models work by estimating the probability of a token or sequence of tokens occurring within a longer sequence of tokens.” (Google)
Machine Learning – “Machine Learning (ML) is the part of AI studying how computer agents can improve their perception, knowledge, thinking, or actions based on experience or data. For this, ML draws from computer science, statistics, psychology, neuroscience, economics and control theory”(Stanford Human-Centered AI)
Prompt Engineering – “Prompt engineering is the practice of designing inputs for AI tools that will produce optimal outputs.”(McKinsey)
References
McKinsey & Company. (2023, April 24). What is ai?. What is AI (Artificial Intelligence)? https://www.mckinsey.com/featured insights/mckinsey explainers/what-is-ai
Manning, Christopher (2020, September). Artificial Intelligence Definitions. AI-Definitions-HAI.pdf https://hai.stanford.edu/ sites/default/files/ 2020-09/AI-Definitions-HAI.pdf
Wired Magazine (2023, February 8). The WIRED Guide to Artificial Intelligence. What Defines Artificial Intelligence. https://www.wired.com/story/
guide-artificial-intelligence/