by Gary Bass
In the 1970s, Arthur C. Clarke famously observed, “Any teacher who could be replaced by a computer, should be.” He was referring to the idea of his ‘electronic tutor’ which would have had easy access to facts and figures and should be used to train teachers. He was of the view that teachers could not be replaced. Meanwhile, 37 years later, that observation still holds true.
Over the period since Arthur C. Clarke made his observation, teaching, learning and schooling have evolved. However, the fundamental role for a learner is about experience, acquiring knowledge and gaining understanding.
Despite rapid advances in computing speed, connectivity and availability, the human teacher continues to provide a superior learning experience for students. Teaching methods may have undergone a multimedia transformation, but the fundamental approach has been largely constant. Telling stories is the way humans relate to knowledge. Basic facts and data are now easily accessed anytime, anywhere. With the advent of 24/7 data availability, a ‘fog’ has descended over the ability to make and take decisive action.
Data-driven decisions are now more difficult than ever before, when perhaps what is needed is information-driven decisions. Means and modes often disguise more than they reveal. When teachers apply their semantic skills, it is the outliers which are of interest. Computer learning obliterates these ‘blips’ and normalises any data to reflect the central seeking tendency of big data.
An arithmetic mean can obscure details; for teachers it is the ‘blips’ which give the insight. (Yau, https://flowingdata.com/2017/07/07/small-summary-stats/)
Teaching and learning is all about personalised experiences. An individual’s perspective is the means of engaging and empowering a learner to seek their own understanding. Schooling with adaptive algorithms may identify a student’s progression point as an historical statement; however, improving that position requires aspiration and inspiration by the learner to shift their achievement to a higher level.
Artificial Intelligence (AI) and computers are brilliant at recording history. Precedent (past precedent is a tautology – future precedent is a guess!) may or may not predict future achievements; however, as the 11 plus exams demonstrated, they do limit future achievement by imposing limited expectations on the learner.
My students who undertook adaptive mathematics testing soon learned that by providing two clearly wrong answers early in the test, they were then provided with an easier workload and gained a better report because they ‘improved’ after an initial low-grade start. If they achieved 100 percent correct, the questions were more difficult and their report showed no improvement over the course.
AI systems can be ‘gamed’. SuccessMaker, in 1997, was an adaptive learning program which received huge hype. The idea was that students would undertake a programmed learning course and emerge with improved knowledge of the topic under study. At enormous cost, this program was deemed to be a failure as it treated all students the same. There was no personalised dimension. Students did not engage, because they were not empowered to take responsibility for their own learning. There was no urge to discover and use the information to make or tell a story. Every input students made was assessed, with no time to ‘play’ and make mistakes. Students quickly discovered that no matter what they did, there was always more to do. (Abbott, 2000)
Project-based learning has received some attention over recent years. While very expensive in time and resources, the results are very impressive. The program can be totally individualised and students negotiate many aspects of their learning. Students become empowered beyond the period of study and often continue with strong interest and activity long after assessments and reports have been issued.
AI has strengths, but the ability to have semantics is not yet one of them. Meanwhile, AI can release teachers from the drudgery of administrative tasks, such as smart notices, where only relevant and appropriate information needs to be presented and everything is voice-activated. Typing is no longer a skill necessary for anyone. Filing systems and database lookups provide just-in-time rather than just-in-case information. In that regard, dramatic culling of study topics can be undertaken. Teachers may dream of the ultimate correcting algorithm; however, there is no substitute for a teacher reading for understanding (or watching a student-made video clip), questioning a student to elicit a response then weighing up and gauging their level of achievement.
For example, fractions as a topic in mathematics could easily be incorporated into an incidental experience rather than dominating mathematics at many year levels for years with repetition and duplication. Similarly, graphing would become data visualisation, logic would be included into early and middle years. Currently, logic is not explicitly studied at any level, including VCE IT. Yet logic is a basic skill in every subject at every level. There are many other topics which dominate student study requirements that are now redundant in the age of ready data access. Many of the reasons why these topics remain is as a legacy or simply because they are easily tested and can be commonly agreed as a ‘standard’. Anything that would be a replacement will be subjective. So, fractions remain.
Taken to the extreme, AI threatens the concept of schooling, where groups of students are arranged in classes and seek sufficient knowledge to be certified or pass at a pre-determined standard. There is no requirement for a student to go anywhere if the AI can be available 24/7 and acceptable progress on learning tasks can be demonstrated by posting online.
The flipped classroom movement has attempted to increase the value of the face-to-face opportunity. Teachers and groups of students in a learning setting provide an opportunity to learn more and faster than is possible online and when physically isolated. Similarly, the tendency to wax lyrical about the engagement potential for virtual reality (VR) experiences overlooks the need to be more than entertaining. As television and videotape had previously claimed, images will revolutionise the way students learn, although that statement depends more on the quality of the content and the task than a clever walk-through in ancient Rome. Being a virtual tourist in places inaccessible by time or space may or may not lead to greater learning or gaining a better understanding.
Teachers provide the insights in such situations, guiding the student experience. Whether AI systems can scale to provide greater exposure to ‘the best’ teachers remains to be seen. Whether this can be automated is also a question yet to be resolved.
If the purpose of education (and schooling) is to better equip learners with wisdom and knowledge to be productive citizens, then the current schooling and AI systems fall dramatically short. Ask anyone about their most memorable moment of schooling and, without exception, it is never the time they gained a high score on an assessment. Invariably, it will be an emotional moment when they gained an insight into themselves or others. That is a part of the human condition. AI can assist, supplement and enhance; however, education, learning and teaching should be so much more than the simple algorithms currently and foreseeably possible. Perhaps re-visit this topic in another 37 years (2054) and see how the progress is going.
- The International Joint Conference on Artificial Intelligence (www.ijcai.org/proceedings/2017/)
- To Know but not Understand: David Weinberger on Science and Big Data (https://www.theatlantic.com/technology/archive/2012/01/to-know-but-not-understand-david-weinberger-on-science-and-big-data/250820/)
- It’s Alive – Artificial Intelligence from the Logic Piano to Killer Robots by Toby Walsh (www.blackincbooks.com.au/books/its-alive)
For a full list of references, email firstname.lastname@example.org
Awarded Victorian IT Leader of the year 2016, Gary Bass has presented at numerous science, IT and mathematics conferences including, most recently, Unreal science VR/AR, Beyond real science: Simulations of collected data using Wolfram Mathematica and SystemModeler, Big Data needs Huge Analysis: Data Visualisation for schools using Tableau and Mathematica. In addition to being an Apple Distinguished Educator, Gary is currently President Mag-Net: Magnificent network: Online STEM Educators Association 2016–17 and a teacher of VCE Informatics at Distance Education Centre Victoria.
Latest posts by Education Technology Solutions (see all)
- A Strategic Implementation of Contemporary Digital Technologies - July 11, 2023
- Why Australian universities need to innovate, invest and transform to remain globally competitive - December 9, 2022
- BenQ Launches World’s First Windows*-Based Smart Projector with Intel Inside for Enterprise - October 14, 2022