Lifelong Personal Teachers The Next Step in Educations Evolution
As artificial intelligence improves, the development of virtual presonalized teachers will become inevitable. Steve Downey, associate professor of education at Valdosta State University, explores the ramifications for educators.
If you asked a student "what traits make for a good teacher", you're likely to hear responses such as "someone who understands me," "someone who knows the content and can put things in terms that make sense to me," and "someone who is accessible when I need help at home, not just in class." In short, they want a personal teacher.
In today's economic, political, and educational environments, the notion of a personal teacher is out of reach for everyone except the wealthiest of the wealthy families. In the next 5-10 years, however, that reality will change. Advances in voice recognition and artificial intelligence over the last 7-8 years have led to the mainstreaming of virtual assistants such as Apple's Siri (released October, 2011), Google's Google Now (June, 2012), Microsoft's Cortana (April, 2014), and Facebook's M (August 2015). Continued advances in artificial intelligence, specifically the branch known as deep learning, combined with expansion of cloud-based storage and computing services will extend the current capacities of these virtual assistants and turn them into lifelong digital companions.
The progression from lifelong digital companions to lifelong personal teachers is only another series of algorithms away. The technology underlying those algorithms already is being developed. In March 2013, the United States Defense Advanced Research Projects Agency announced a new initiative to "decisively reduce the current barriers to machine learning and foster a boom in innovation, productivity and effectiveness." In January 2014, Google paid $400 million to acquire London-based DeepMind Technologies, a small startup focused on deep learning research. Prior to this acquisition, Google's director of research, Peter Novig, acknowledged that Google already employed more than five percent of world's leading minds in machine learning. Google isn't the only tech firm investing heavily into machine learning. At the same time Google was buying DeepMind Technologies, IBM announced it was spending $1 billion to create a new business unit for its Watson machine learning technologies. Microsoft, Apple, Amazon, Facebook, and numerous other big money organizations are active in the field as well.
In addition to their work in AI-based personal assistants, Apple and Microsoft have long been active in education. Similarly, Google launched Google for Education in November 2010. In 2014, IBM announced Watson for Everyone, intended to bring AI-based analytics to individuals. Even Facebook has expanded its initial tentative entry into education with its recent (September 3, 2015) announcement of the Summit K-12 Education Project, which is envisioned to integrate a variety of learning technologies into a single interface and enable students to create "personalized learning plans" articulating their own learning curriculum. It will be a small step for these and other tech firms to turn their AI-powered sites on classroom instruction.
The progression of technology is inevitable and undeniable. The question is, how will education evolve as a result of it?
Until the Web came along, education had taken few radical evolutionary steps since Gutenberg's printing press. Granted, there are a lot more schools now than in 1440. However, within these walls we still matriculate students based upon imprecise units of credits. We still, too often, arrange our classrooms into rows; and we still, too often, have all of the students learning at the same pace regardless of prior knowledge or intellectual ability.
With the arrival of AI-based lifelong personal teachers, however, multiple aspects of education will be forced into reform. Classroom teachers no longer will be the primary source of knowledge. In the future, when a student needs something explained, the student won't have to wait for class; the student will simply say to his or her personal teacher, "Siri, I don't understand the chain rule. Please explain it to me." Siri will query its archives, apply the appropriate AI algorithms to accommodate the student's existing knowledge and instructional preferences, and then explain the chain rule to the student in terms the student can understand and internalize. Consequently, today's teachers will need to evolve their instructional focus and delivery methods.
Looking beyond the classroom, K-12 schools will need to reform how they manage their curriculum and the manner in which students progress through their programs. Facebook's aforementioned Summit project already is targeting this aspect of K-12 programs with its Personalized Learning Plans. In addition, competition quickly will arise from other tech firms and start-ups offering cloud-based archives and downloadable algorithms for courses and entire instructional programs customizable to individual students.
Policymakers' Role in Enabling Education's Evolution
Before any sweeping reforms in education can occur, state and federal policymakers must come to grips with the implications and scope that artificial intelligence will have on education — in and out of the classroom. To date, however, it appears policymakers aren't even aware of the change heading towards them. Searches of the U.S. Department of Education's web site as well as the sites of the four regional consortiums that provide policy making guidance to state legislators (i.e., New England Board of Higher Education, Southern Regional Education Board, Midwestern Higher Education Compact, and the Western Interstate Commission for Higher Education) finds few mentions of artificial intelligence and no clear cut policies or recommendations for legislatures to debate, much less adopt.
Being a highly regulated industry, teachers and school administrators are limited in how fast they can modify their day-to-day operations. This legislated inadaptability even further heightens the need for educators to work with industry experts and legislative officials to establish guidelines for classroom and general purpose uses for AI as well as funding and collaborative endeavors to help ensure students and schools with limited financial resources aren't left out of the change process.
Have faith Chicken Little, the sky isn't completely falling.
With change comes opportunity. Just as occurred with the advent of online learning, new forms of learning and program operation will arise with the arrival of lifelong personal teachers. Instead of lecturing and the memorization of facts for unit exams, classroom teachers can incorporate new resources, forms of reasoning and collaboration. Instead of forcing students to complete a predefined sequence of courses regardless of their existing abilities, institutions will learn to support personalized learning plans that culminate in the achievement of defined skills and knowledge. Numerous new lines of research will emerge, e.g., examining the duality of concurrent machine learning and human learning, the longitudinal use of AI-based personal agents to help individuals with problems they often hide from others — victims of bullying, severe depression, etc.
The exact form and scope of these endeavors remains to be seen. What is clear, however, is that today's schools need to shake centuries of bad habits and learn to adapt and evolve quickly. New models need to be developed for teaching, retaining, matriculating and graduating students. If they aren't, Silicon Valley and tech firms around the world will impose their collective wills in the form of innovative new programs of learning. This leads to one final question, "Siri, how can my school evolve?"