Artificial intelligence (AI) has been a topic of controversy within academia, particularly over the past decade, with the emergence of ChatGPT and other AI software, such as Mathway (https://www.mathway.com). At the University of Wisconsin- Green Bay, an AI policy has been established for all students and professors. The university permits the use of AI software that is not integrated within the University of Wisconsin’s SIS or Canvas networks, provided the student obtains approval from the collaborating professor in the course where the AI software is being used. Professors typically provide a section in their syllabi or policies, advising students on which types of AI are accepted in their class. All AI usage by students must be reported to their current professor(s) for approval if it is not listed on the professor’s accepted AI programs list or if it is not on the UWGB Library list of accepted AI programs. The University of Wisconsin-Green Bay’s list of accepted AI programs can be found here: https://libguides.uwgb.edu/CiteYourSource/AI. The impact of AI in the classroom is often described as a “mixed bag,” with varying opinions and outcomes. As AI is used more frequently, questions arise regarding whether AI assists students or facilitates plagiarism. Several factors contribute to this debate, including the type of AI program used, its purpose, and whether it influences the outcomes of a professor’s instruction.
Let’s Get to Know AI
There are three types of AI commonly utilized:
• Narrow or weak AI – AI designed for specialized tasks.
• General or strong AI – AI developed to mirror human cognitive abilities.
• Superintelligent AI – A speculative concept envisioning AI surpassing human intelligence.AI – a futuristic notion that envisions AI surpassing human intelligence.
ANI (also known as unsupervised AI) is AI used for mundane tasks, such as locating information relating to a specific task. This form of AI is generally used to cross reference databases, or to have use in a one-to-one function when retrieving outcomes to a specified request.
General AI (also known as supervised AI) is used to mimic a human’s thought process when drawing conclusions. This AI’s algorithms are still entered by a human into the programming language used by the AI, through a software tool called a syntax (a tool within a software program that is used to convert programming language into machine language so that the AI software can understand what it is being instructed to perform). Even though the algorithmic process itself is not supervised, humans still monitor the specific AI results once compiled.
Superintelligent AI (also known as unsupervised AI) is generally within use at Universities by Ph.D.s in creating algorithmic thinking patterns that involve long time absence from monitoring the results of AI, for self-learning purposes. This type of AI is commonly known as either Neural or Deep-Learning Networks. There is usually more than just computer science at work here. It is known that neurologists, computer scientists, and psychologists all play a part in this AI’s development.
The ongoing debate surrounding AI in education primarily involves narrow and general AI, which are available to the public and commonly used in classrooms today.
A History Lesson in AI
The concept of AI has existed for a considerable time. Major universities, such as MIT, began using Metareasoning theory to allow software to recognize complex matters (METAREASONING: Thinking About Thinking. Michael T. Cox. Anita Raja. MIT Press, 2011). Metareasoning is the coding theory that AI can or cannot develop the ability to “think about thinking.” The human brain unconsciously processes thoughts when using one of the five senses to recognize a person, place, or thing. For example, when someone sees a stove burner, they unconsciously evaluate whether it is hot by considering factors like the burner’s color or the presence of heat.
The challenge is to implement this Metareasoning theory into software recognition programs, such as Neural or Deep Learning Networks. Another issue is whether machines can learn from mistakes and draw conclusions—this remains a significant concern in current AI usage.
AI development advanced with the introduction of Generative Adversarial Networks (GANs), which recognized and labeled images, allowing AI to distinguish real objects from imitated ones, such as in facial recognition technology. Initially considered a noncritical task, AI’s capability in this area has significantly improved over time. By 2016, AI audio generators like WaveAudio were publicly recognized for their ability to convert audio to text.
Today, a wide variety of AI software platforms exist, each serving different functions. This opens up debate on theories like [P vs. NP] (The Golden Ticket: P, NP, and the Search for the Impossible. Lance Fortnow. Princeton University Press, 2017), which suggests that every algorithmic problem has a solution. A one-million-dollar reward is offered by the Clay Institute of Mathematics for solving this theory. As AI becomes more integrated into everyday life, academia continues to debate the safety and educational quality of AI usage by students.
AI and Public Opinion
Interviewee: Ben Khoun – Library Manager
In an educational interview with library manager Ben Khoun, it was revealed that AI has become so integrated into society that Khoun was unaware the library’s Enterprise program is AI software. Khoun explained that Enterprise is responsible for locating books within Wisconsin’s library system, estimating their arrival time when requested, and determining their checkout statuses, identified by unique ISBN numbers.
Based on knowledge from UWGB’s Database Management course, AI is involved in the Enterprise software. It calculates the exact location of books, estimates arrival times when requested from other branches, and determines checkout status and return dates. AI is responsible for gathering information from multiple database tables to provide results. Khoun expressed surprise upon learning about the extent of AI’s use in the library system. However, he also disclosed that a similar system is used for book checkouts, but he could not reveal details due to the presence of personal information.
Khoun acknowledged that AI has been extremely helpful in his line of work. However, when asked about its use in schools and universities, he expressed uncertainty. This sentiment reflects a common reaction in interviews and surveys, where many individuals are unaware of AI’s extensive use in everyday systems.
The Big Picture:
There remains significant debate about AI’s impact on student learning. Some argue that AI may detract from the quality of education, as students using AI tools like chat-to-text may not fully engage in writing or formatting tasks required by professors. However, others recognize that AI provides benefits, such as aiding students with language barriers or offering feedback on drafts. AI policies differ from university to university, with varying views on its role in education.
