Associate Professor Jack Wang, a teaching-focused microbiologist at the University of Queensland, was both impressed and worried when he first encountered the AI-powered chatbot ChatGPT. Was he witnessing a future in which AI would replace him as a teacher?
In response, he produced the video “Can AI Replace Professors?” and posed a series of questions to ChatGPT to assess whether it would be possible to successfully replace teachers. “It was an interesting experience making that video because I had to challenge all of my preconceived notions about what my value as a teacher actually is,” Wang, who was selected Australian University Teacher of the Year in 2020, told Panopto.
ChatGPT, which was released on November 30, 2022 by OpenAI, broke the internet, amassing over a million members in only five days. It was straining to keep up with demand two weeks after its introduction; it had grown so popular that user capacity had been reached. Not surprising given what it has to offer. ChatGPT is not the first AI-powered writing tool to be made available to educators. However, it is the first to achieve human-like conversational and writing abilities. And that’s not all it can do; it can also write programming, design websites, and compose music.
Chatbots can answer students’ questions about class scheduling or check in with them about their mental health. AI-generated emails can remind students about important deadlines, prompt them to register for classes, turn in assignments and pay their fees on time. And, in a particularly controversial use, AI-based software is increasingly able to detect plagiarized assignments. One professor at Georgia Tech, even used AI to build a virtual teaching assistant, called Jill Watson. Turns out that “Jill” receives very positive student evaluations.
AI helps alleviate the stress of menial activities on faculty and staff
It is common knowledge that teachers are under-staffed and overworked. A 2022 report indicated that 75% of teachers face ongoing job-related stress, while 59% are burned out, an issue brought to light by the COVID-19 epidemic. Burnout causes instructors to leave their positions, resulting in a countrywide shortage of teachers in the aftermath of the epidemic – a lack of 2.6 million educators who have resigned, to be exact.
If you’re under-resourced and overworked, automating menial jobs could be the solution. AI may be used to automate many elements of education, including data entry and paper-based grading, allowing teachers to focus on more vital activities. According to the McKinsey Global Institute’s 2018 research on the future of work, existing technology may automate 20-40% of the hours spent by teachers on activities. Teachers can save time by using Panopto’s smart search instead of transcribing, labelling, and categorizing all of their video content. Automatic voice recognition (ASR) collects and indexes every spoken word, while optical character recognition (OSR) indexes every word that appears visually on-screen using AI-powered technology.
AI can screen potential students for admission
According to Eric Wang, senior director of AI at Turnitin, a tool many institutions use to check for plagiarism, much of the AI-powered software used by top colleges and universities is confined to very routine tasks such as enhancing back-office operations. Where it starts to get worrying, though, is when AI becomes influential in much more important, higher-stake decisions.
For example, AI can be used to predict how well students will perform if admitted to a certain college and can aid in determining their financial need. Hundreds of colleges pay for private platforms that conduct extensive data analysis on previous classes and use it to score applicants for admission on factors such as likelihood of enrollment, amount of financial aid required, likelihood of graduation, and likelihood of being engaged alumni.
Some institutions even utilize AI to forecast applicants’ performance based on how they engage with a school’s website and respond to its messaging, which the service provider claims is 20 times more predictive than depending solely on demographics. New York University, Southeast Missouri State University, and other institutions have all adopted a tool that does this very thing, showing that many top colleges are using AI to analyze potential students.
Many students now receive messages from organizations such as AdmitHub, which promises a customizable chatbot and text messaging platform dubbed “conversational AI” to “nudge” approved applicants into putting down deposits. According to the company, it has reached more than 3 million students in this manner on behalf of hundreds of university and college clients.
Georgia State University has been at the forefront of adopting chatbot technology in higher education. Their chatbot, Pounce, has been active since 2016, providing hundreds of thousands of responses to inquiries from prospective students. As a result, the university has experienced a 20% decrease in “summer melt” – the phenomenon where students enroll during the spring but fail to attend in the fall.
Additionally, Georgia State was among the pioneers in creating cost-effective, always-available AI teaching assistants. Their AI assistant, Jill Watson, has been so effective that some students were unable to distinguish between interacting with the AI and a human teaching assistant during their studies. This highlights the significant advancements in AI technology and its potential in enhancing the higher education experience.
AI can assist when students are already admitted as well
As the integration of AI in higher education becomes more prevalent, universities around the world are developing AI-powered teaching assistants to support students in various ways. These digital assistants not only provide instant assistance but also help students manage their time and resources more effectively.
At Staffordshire University in England, students have access to a digital companion called Beacon, an AI-driven teaching assistant that offers personalized reading suggestions and connects students with tutors when needed. Similarly, Deakin University in Australia has developed an AI assistant named Genie, which possesses knowledge of a student’s engagement with online course materials. Genie can also monitor students’ locations and activities to provide timely reminders, such as nudging students to leave the dining hall if they’ve spent excessive time there or even encouraging them to visit the library. These AI-based innovations are transforming the higher education landscape by offering tailored, real-time support to students.
As the number of students in online classes grows, many colleges are increasingly turning to AI to help with grading, particularly as the pandemic has accelerated the shift to remote learning. For instance, even before the pandemic, Southern New Hampshire University (SNHU) was exploring ways to leverage AI for efficient grading of its predominantly online student population, which amounts to nearly 150,000 students, according to Faby Gagne, the executive director of the university’s research and development division. Moreover, SNHU is beginning to utilize AI not only for grading but also for teaching purposes. Gagne’s team has been experimenting with AI systems that analyze various aspects of student engagement, such as speech, movement, or response time during video lessons. This information is then used to evaluate student achievement and provide a more personalized learning experience.
Turnitin, a company primarily known for its plagiarism detection services, also offers AI-based language comprehension products for evaluating subjective written assignments. One such tool can categorize written tasks, enabling educators to address a common error or provide guidance only once, rather than repeatedly identifying, commenting on, and grading the same mistake. Although instructors still verify the accuracy of the AI’s assessments, the reduction of repetitive tasks allows them to devote more time to teaching and engaging with their students.
AI is making decisions previously made by faculty, but can AI be wrong?
AI-based solutions are increasingly being offered to higher education institutions to assist in decision-making processes traditionally handled by faculty members. For instance, ElevateU employs AI to examine student data and provide customized learning materials to students based on their responses to questions. The system tailors the learning experience by selecting the most suitable format for each student, such as providing a video lesson instead of a written one if it deems that the student would benefit more from that approach.
However, research indicates that AI tools can sometimes produce incorrect results or even be manipulated. For instance, a group of MIT researchers developed a computer-generated essay that, while essentially nonsensical, included all the elements an AI essay evaluator looks for. The AI system awarded the nonsensical text a high score.
In another case, an AI chatbot named Lola in Spain answered over 38,700 student inquiries with a 91.7 percent accuracy rate. This means it provided at least 3,200 incorrect or incomplete responses. Jarrod Morgan, the founder and chief strategy officer at ProctorU, a company that manages and monitors online exams for educational institutions, stated that “AI alone is not a good judge of human behavior or intention.” He emphasized that humans generally outperform machines when it comes to understanding and assessing human actions and motivations.