Educators are at the ends of their ropes. So suggests a fall 2020 survey by RAND Corp., which found that a quarter of all teachers were thinking about leaving education.
Remote learning and COVID-19 are partly to blame: More than half (57 percent) of teachers said they worked more hours per week during the pandemic than they did before it, according to RAND, and 80 percent reported feelings of burnout as a result. Even before COVID-19, however, former public-school teachers were struggling, and reported finding better pay, better work/life balance, more resources and a more manageable workload in jobs outside of education.
“Part of the problem is that teachers spend a lot of time doing things that … in their view are not the best and highest use of their time,” says former teacher Jake Bryant, now a partner at management consulting firm McKinsey & Co., where he serves the company’s education practice. “Nobody becomes a third-grade teacher because they love collecting permission slips and filling out attendance sheets. What motivates you to get into the profession is interacting and engaging with students, and helping them learn.”
Technology might be the remedy, according to Bryant, who cites McKinsey research finding that teachers currently spend as much as 40 percent of their time on activities that could be automated. That’s hours every week that could be reallocated from administration to education.
To be clear, robots won’t replace teachers. By deploying artificial intelligence (AI) in the classroom, however, experts like Bryant believe schools can remake education in ways that make teachers happier and students smarter.
Because teachers are stretched so thin, AI can make a significant difference by streamlining administrative tasks.
“The strongest use cases for AI in education are those that automate the more tedious aspects of the educator workload,” says Eric Wang, senior director at education technology company Turnitin AI. Take grading, for instance: Turnitin’s Gradescope platform uses AI to analyze students’ work. It finds and groups similar answers, then creates an automated rubric that helps teachers grade assignments in half the time.
“The advantage of artificial intelligence is its ability to scale complex tasks or reduce time spent on redundant tasks,” Wang continues. “AI-powered learning products give faculty more time to focus on teaching and the interpersonal aspects of their interactions with students.”
In 2016, Georgia Institute of Technology computer science professor Ashok Goel used AI to create a time-saving solution of his own: a virtual teaching assistant named Jill Watson that’s available 24/7 to answer routine questions posed by graduate students, such as when exams will take place.
“Having an AI agent available to answer basic questions allows students to get help any time, any place, and reduces the load on teachers,” Goel explains. “Teachers spend so much time on mundane activities. Freeing them up would allow them to engage more deeply with learners, which is what teachers should be doing in the first place.”
The Power of Personalization
No matter how much of it they have, AI can maximize teachers’ time by way of personalized and adaptive learning.“The typical classroom is incredibly diverse in terms of where students are starting from, what things they understand, what motivates them and how they feel engaged,” Bryant says. “Technology creates an opportunity for teachers to give instruction that’s more personalized with less effort.”
Thinkster Math builds a customized learning plan for each student that includes worksheets and video tutorials. THINKSTER MATH
Knewton’s Alta and Thinkster Math are two examples of adaptive learning. The former is an adaptive courseware system that uses AI to create personalized learning experiences in college-level math, science and economics. As students complete assignments inside the system, AI automatically recognizes knowledge gaps and retrieves content to address them. Students who understand the material can breeze through and move on, while those who don’t will receive extra, targeted instruction. All the while, the system feeds data to instructors to inform subsequent instruction.
“The technology accelerates students through pieces where they don’t need help and spends more time with them in areas where they do,” explains Matthew Leavy, executive vice president and general manager of education publishing at Wiley, Knewton’s parent company. “And for instructors looking over large classrooms, it helps them understand where they should focus their teaching and where they don’t need to spend as much time.”
Thinkster Math is a tutoring platform for K-12 students. First, learners take an online assessment that determines their knowledge in a given domain. Then, the system builds a personalized learning plan and assigns custom worksheets to complete. AI captures students’ work, provides video tutorials when learners get stuck and isolates learning opportunities that human instructors can focus on during subsequent live tutoring sessions.
“We’ve married man with machine,” says Thinkster Math founder and CEO Raj Valli, who likens the technology to a swimming coach who watches his swimmers’ every stroke. “If you tell me to jump in the pool and swim back and forth, I’m never going to be a good swimmer. But if you jump in the pool with me and point out that I’m not kicking my right leg or using my left arm, then you can make me better. That’s the kind of observations our tutors are able to make using our technology.”
Reading Progress, a brand-new tool from Microsoft, applies AI and speech recognition to reading fluency. Teachers assign reading level-appropriate passages that students read on camera and submit for review; with the help of AI, instructors can then assess performance and identify reading errors.
“Everything Reading Progress does, a teacher could do if they were able to sit next to each of their students all day and coach them on every single word they read. But in a classroom of 20 or 30 students that’s often not possible,” says Anthony Salcito, Microsoft vice president of worldwide education. “Reading Progress helps teachers quickly assess how students are progressing so they can coach them where they need help.”
The Gradescope platform uses AI to read and digitize a student’s handwriting.
AI in the future might optimize not only individual curriculums, but also entire classrooms. For example, Goel says AI could be used for “matchmaking” — pairing students with the teachers and schools that are best suited to them based on their learning style.
Meanwhile, Sean Ryan, president of the School Group at McGraw-Hill, says there’s an opportunity to organize students into classes based on aptitude instead of age. “For the most part, today we sort students chronologically no matter what. But AI gives us the ability to group students based on what they’re ready to learn next,” says Ryan.
McGraw-Hill’s ALEKS adaptive learning program uses AI to create personalized learning paths for students in kindergarten through college. “That can be hard to embrace because of social components. But with more education taking place in hybrid and online environments, there’s no reason not to put an eighth grader in a pre-calculus class if they’re cognitively ready,” Ryan says.
It’s the beginning of a new era wherein learning is a journey instead of a destination. That makes teachers navigators — which is precisely what most of them want to be.
“Teachers become teachers to help children maximize their potential,” Ryan concludes. “By allowing them to focus more on the social components of learning, technology helps them have the kind of impact they got into the profession to have.”