Do Students Dream of Electric Teachers?

A.I.’s Possible Roles in Education

Do+Students+Dream+of+Electric+Teachers%3F

Jayden Tutza, staff writer

Artificial intelligence—or A.I.—is a medium believed to be “the future of learning” in the eyes of many researchers; so much so, in fact, that panels about A.I.’s impacts on education have been gathered to analyze its possible effects as recently as November of 2020. Take Apple’s Siri, Microsoft’s Cortana, or Amazon’s Alexa, A.I. assistants that respond to human inputs. Programs similar to those could become standard in classrooms, following the predicted growth of the technology. This doesn’t mean that robots will be replacing teachers any time soon—not in the opinions of researchers and educators alike.  

To truly grasp A.I.’s place in the classroom, one must first understand the basics of A.I. According to Lisa Plitnichenko, writer of “5 Main Roles of Artificial Intelligence in Education,” A.I. is split into four specific categories: assisted intelligence, which requires help to make decisions; automation, which performs repetitive tasks; augmented intelligence, which has the capability to make human decisions; and autonomous intelligence, which has the ability to adapt based on different stimuli without human assistance. Assisted intelligence and automations are unable to “learn” like the other two forms of A.I., instilling them with an adeptness at common, often tedious, tasks. Plitnichenko goes on to explain that “A.I. tools must comply with three basic principles: learning, [the process of] acquiring and processing new experience, [thus] creating new behavior models; self-correction, [the process of] refining the algorithms to ensure the most accurate results; [and] reasoning, [the process of] picking up the specific algorithms to resolve a specific task.” A.I., through analysis of patterns, forms connections like those of the human brain; in fact, A.I.’s learning resembles that of a human infant, with similar “thought” patterns that quickly develop as conclusions are drawn. According to an article entitled, “What Can A.I. Learn from Human Intelligence?” A.I. is developed “using concepts from everything from developmental psychology to animal behavior. For example, DeepMind is currently training AI neural networks using cutting-edge understanding of dopamine-based reinforcement learning in humans. ‘We’re helping AI systems make better predictions based on what we’ve learned about the brain,’ [Matthew] Botvinick says.”  

So, what does this mean in terms of education? A.I., alongside its ability to perform simple tasks such as grading or paper sorting, has the capability to tailor instruction to each student based on how they learn best. Ms. Kasie Windfelder, the principal of Navarre High School, describes this process as “a very laser-targeted level of differentiation that would be so specific that a human… may not be able to get to that nth degree.”  

A.I. is already being applied for younger students—including those attending Holley-Navarre Primary School—in the form of Accelerated Reading, a program that analyzes a student’s pace and knowledge through A.I.-assigned reading quartiles. “The questions are appropriate for where they’re at,” continues Windfelder, “…and push[es] them just slightly so they can keep growing.”  

So, what would happen if that same programming idea were taken a step further, personalizing entire lessons to each student? A.I. still has a long way to go, according to an expert panel about the future of education under A.I. The conclusions garnered from the panel, given in a report entitled, “A.I. and the Future of Learning: Expert Panel Report,” point to a large conflict between A.I.’s capabilities and its tendency to fail ungracefully. Limitations of A.I. include a lack of awareness to learning differences, a lack of ethics, and a limited interface. Conclusively, A.I. does not have emotions or a conscience, something that “only a teacher brings to the table,” says Windfelder. “With teachers, the first things that they focus on in classrooms are relationships with students.” With these limitations in place, the “self-aware A.I.” that is often demonized in video games and television—take your constantly-evolving robot or your dating-sim girlfriend with control issues—is unlikely to occur, due to the narrowness of A.I.’s current programming platforms. 

However, even with these limitations, A.I. is growing faster than ever. No, an emotionless A.I. isn’t going to brainwash school students—in fact, its assistance has already made an impact on students and teachers alike. “It helps with the efficiency of teachers,” proposes Windfelder. “An example would be when a teacher has a portion of their test where they use multiple choice, and then [they] can use a Scantron… [to] quickly grade the test. It seems like a lot of people like that instant gratification, that instant feedback; and so, you’re able to see at least a portion of your grade right away.” Experts, too, are still willing to explore the future of A.I. on education; A.I. has an inherent skill of “analyzing learner performances in collaborative groups, simulations, and other rich forms of knowing, [alongside] …making invisible aspects of teaching and learning more visible, such as uncovering missed connections between different skills… to deepen support for learning across the related but separate contexts.” 

A.I.’s role in the classroom, at least until it begins to develop an idea of mindfulness, may be to supplement teachers and students instead of completely taking on the role of an educator. The decision of A.I. implementation isn’t as simple as a red pill or a blue pill; it requires an analysis of risk that may be years in the making. Even if these advanced technologies are employed, they will require at least minimal human input and assistance in order to function, needing human moderators to ensure the safety of students and the A.I. “I think there is some value in it,” concludes Windfelder. “It’s definitely not going away. I think it’s something that we—all of us in education—should be conscientious of. We need to make choices that are in the best interests of the students, versus what the new fad is with technology.” 

Sources: 

CIRCLS AI Report Nov2020 

What Can AI Learn from Human Intelligence? (stanford.edu) 

What does AI in education look like? Here’s what research shows. (hechingerreport.org) 

5 Main Roles Of Artificial Intelligence In Education – eLearning Industry