Machine learning uses data to predict what you will want to eat, consume, purchase and more. So why not for learning?
The internet looks nothing like it did when it started, but schools have taken a slower approach to change!
No one involved in the early construction of a network of networks saw what we have now coming. Therefore how can we know what comes next, but indulge me in an attempt at envisaging the potential impact of modern internet capabilities on the institutions of learning? Might the ever evolving internet have the potential to finally nudge education to innovate and evolve?
“From the moment you wake up, the web is trying to anticipate your intentions. Since your routines are noted, the web is attempting to get ahead of your actions, to deliver an answer almost before you ask a question.”
Understanding the 12 Technological Forces That Will Shape Our Future by Kevin Kelly
The extract from my most recent Kindle acquisition prompted me to apply this idea to my passion, education. Machine learning uses data to predict what you will want to eat, consume, purchase and more. So what will this look like for learning? Google Drive already makes a stab at which documents you want to access when you open it. Google Sheets are starting to predict the formula you want to use, and it will only get better at it. Get your hands on the new Jamboard app which turns your sketch of common items into a neat icon. How? Using all the drawings people have added to quickdraw. If the data is there, a machine can learn from it. Learning is a very complex system but our use of data to measure learning is extremely simplistic.
Learning machines
If you live in Scotland as a 14 year old, this information alone gives a reasonable indication of what you are studying. Furthermore, your documents and data will too. Is it not inevitable that you will receive study materials, curriculum updates, pop quizzes and feedback based on this basic information? Having studied Pythagoras three weeks previous you can be provided with review materials, your notes and suitable problems to solve at the optimal time to review it and ensure it begins to embed in your long-term memory. This is available information ‘the internet’ can use to support your studies with timely content. This is not Sci-Fi, this is functionality many of us experience through our devices because we have traded access to our data for the tools to make our lives a bit easier or more productive. If this kind of automation and feedback were available to students, the kind of efficiency I certainly didn’t achieve (ask my students!), why would they need their teachers? A question you only ask if you are not able to change your view of what a teacher is or could be.
News corporations could not envisage the internet delivering news because their schema did not allow it. They were not able to foresee or imagine all their passive readers and viewers becoming the creators of content, whether that be videos, blogs, social media posts or reviews on Amazon. They assumed the content on the internet would have to be created by them, but that would not be economically viable. Remember, we are abysmal at predicting the future, yet here I am trying!?
What is school?
If a student is receiving feedback on their work quickly and efficiently without having to enter school building then what is the school for? A social place to share your learning, remain mentally sound by providing interaction with other children and experts? Will your schedule be dynamic and each day adjust to your needs? Let’s not forget we like structure… Could a machine learn the structures we work best in?
As a teacher would I have your essay appear in the morning and a meeting scheduled for the afternoon. The marking of your essay against a rubric will already be done the moment you ‘submit’ and I see it too. The thousands of teacher marked essays and millions more marked by machines means the assessment is now better conducted by machine than by a human. Our scheduled meeting is to review and discuss the next steps. Other students may join our tutorial by video link, they are visiting an educational site elsewhere but would benefit from the tutorial so they are added to the meeting if convenient, another institution may have a more convenient essay review of course.
All the while their online professional portfolio is being created and made available to employers and recruiters who might benefit from their skills. Employer data on the attributes of their most productive teams use this to match a students potential profile. This is similar to current sporting recruitment where young athletes with a lung capacity above the mean average might be pushed towards rowing or cycling. As long as every student can participate and have their information accessible in this way social mobility can be accelerated as the screening of portfolios is automated, bypassing our bias for names and backgrounds that we know or are similar to our own.
Don’t be put off by student data being accessed. This can easily sit within data protection laws if the service we want is for students to get prompt and useful feedback, then the data processor is within their rights to use the data they need to provide the service. The data controller, would that be the institution… Or could it be the student’s family? Whether you have “Alexa”, “Siri”, “Cortana” or just “Hey Google..” ringing around your house, you will be familiar with Artificial Intelligence and Machine Learning. My phone gives me driving times to different destinations on different days, based on my previous journeys. So let’s take this scenario to learning. Based on your habits and working patterns might we have these types of prompts from our devices?
“It looks like you are studying, how long would you like notifications turned off”
“This is usually your most productive study time, shall I open your current assignments?”
“Looks like you have finished your essay, would you like to get feedback on it now, if so a lecturer is available in two hours to review the feedback, shall I schedule a tutorial? 7 other students on your course will be present.”
“You are not studying effectively, take a break I will check if your friends are free”
“Ready for a pop quiz on the civil war?”
These interactions could be the way we manage our studies, automated and based on our data. Machines using our responses to improve predictions and support.
Learning Institutions
Where might learning happen and what of the experts, coaches or teachers supporting that process? It would appear unlikely we would do away with physical space dedicated to learning. Our access to online learning is now vast, with reputable institutions offering courses online. However, the completion rates are very poor and these Massive Open Online Courses (MOOCs) may not be the future of learning.
The screen in the Principal’s office provides an interactive infographic of feedback metrics on student participation and progress in courses, student ratings for tutorials and dynamic course numbers. Permanent staff include trained councillors, learning facilitators but the teaching staff are fluid depending on need from week to week. Through their profile, many join via video link to support a burst of course interest or required expertise. Others are longer term on site having had superb feedback and progress metrics. They have built strong relationships and reputations for your institution and also contribute content to the courses. Their involvement with other institutions benefits the cross-pollination of innovations in learning. Why not have Elon Musk as your visiting transportation lecturer?
What are your qualifications?
There are signals of the waning value of traditional academic achievements. Will the transition between learning and work blur? How restricted are we in our exam systems? I am going to go and read more on assessment, the skills needed in a changing world and train some teachers on using G Suite for learning.
Writing this post is foolish as I am planting an artefact online that will surely be embarrassingly inaccurate. However, I will continue to consider what learning may look like as the world changes. I am confident change will come, but in education, I cannot predict whether it will be a systematic evolution or a learner revolution.
Thanks for reading.
Evolving Edtech 