How AI & Chatbots Are Changing Education Worldwide

 Is the FHE teaching capability and capacity improving as fast as it should?

Some of our knowledge about teaching and learning go back to Greek times and still hold true. But that is not to say that more recent research and technology should be ignored.

It is generally accepted that Moore’s Law is about the doubling of digital capacity every couple of years (the original Law actually talked about transistor capacity). Although not directly linked to the teaching, the curriculum or success rates, there are perhaps a few lessons that can be taken from the way that improvements and growth in the digital world impacts our ability to make teaching more effective.

Learning has never been just a classroom based activity and today we have even more learning taking place at work or in the home. This fact also recognises the difference between micro and macro learning.

Early in a career, and often at other times, we all need macro learning. These are the certificate, diploma and degree courses of old. But interspersed between these we need micro learning. Short learning episodes that might last anything from a few minutes to a few days. These are the times when we Google things like “how to do [something] in Word. Or they could be an updating session on the latest software or new machine we’ve purchased. They are so frequent we often neglect to recognise them as learning sessions.

Formal FE tends to ignore these opportunities to become involved in teaching.  But informal FE providers are often very involved. For example informal providers might produce a short but very full course and sell it online. Or it might be a free YouTube video funded by advertising revenue, but informal providers fill these niches.

Despite being micro courses the revenues earned are huge and can run into £/$millions per course. UK formal FE would benefit from paying attention to these markets. The informal providers are often people working from home, often young people, they don’t appear on official registers and are a hidden part of the training economy.  

In this article I want to explore how universities and colleges worldwide are blending tradition with contemporary methods to improve learning. We know many young people are glued to their mobiles and consume huge amounts of content via video. So I’m adding videos to this article for my younger (and more technically able) readers! The twist in the tail is that the videos have been made by some of the technology I’ll be discussing. All videos in this article have been made by a VideoBot based on the article content.

Let me start with an introduction about the widespread use of the automated systems that our student body consume in such huge quantities … and a bit about how automated essay marking is being used in some institutions.

Automated Systems Ubiquity & Automated Essay Marking

Today we are able to buy anything online. From our weekly supermarket shop to music, video, clothing and holidays. We can order, pay and arrange delivery online. Millions of households worldwide now have an Amazon Echo or similar device, speak to Siri on their phone or Cortana via their PC.

Automated Process Systems, or bots, are everywhere. They are so common that we now take them for granted in many situations. Providers use them on their websites and phone systems without even realising that what they are using is a bot.

So why don’t we use them more in the classroom? With time at a premium we don’t we see more use of things like automated essay marking?

Some providers use automated essay marking in combination with peer and teacher marking. It isn’t suitable in all situations. However used wisely, maybe as part of a software self directed package, it has a lot to commend it.

Digital Course Assessments & Student Ratings

What do students think of the teaching they are getting? With students now being considered more and more as customers, their assessment of their courses and teaching is now vital.

For years we’ve used paper systems to undertake evaluations. Are these effective enough? Many would argue not. They would point to the digital system that can be used on the grounds of ease, accuracy and the automatic analysis of results.

They’d further point out that a blend of qualitative and quantitative results can be easily obtained and that these can be viewed alongside other data e.g. attendance, success rates, self assessment etc. and that results can be correlated in this way. Using these methods it is possible to make international comparisons of teaching methods and results.

The Spacing Effect

It is a long established fact that we forget information over time. We know that the use of regular reminders built into any tuition package increases retention.

This isn’t a new insight into learning. It was mentioned by Hermann Ebbinghaus in 1885 (Ebbinghaus Forgetting Curve), and no doubt by others before him.

Good teachers naturally build repetition into their lesson plans; it’s part of what makes them good. The only problem is that students are individuals and exhibit different rates of degradation in their recall. So what if each student could be continually tested as they are taught, and as the subtle signs of memory loss are noticed, they are given additional nudges to aid reinforcement?

For a tutor to do this on an individual basis is next to impossible. But programmes such as  SuperMemo do this as they teach languages and other courses. It is so subtle that every student gets an experience tailored to their needs and learning is accelerated.

Other vendors of Spacing Effect software include Axonify, Grovo, Practice and Edcast, some of which allow providers to develop their own courses that utilise the Spacing Effect.

Student Centred Feedback

Closely aligned to the Spacing Effect is Student Centred Feedback. Where an individual learns at an individual rate and may exhibit preferences to particular learning styles some programmes now use AI and machine learning to adapt to the needs of the individual with tailored tuition.

The Educational Endowment Foundation in the UK has say learning styles are “Low impact for very low cost, based on limited evidence”. However, for many years the impact of learning styles has been considered a fact, and many of us would consider ourselves as having visual, auditory or kinesthetic leanings.  

It could be that AI will cast more light on this and define a more granular learning profile than the present three preferences.

In my own mind I think it worth trying to learn more by using the feedback that results from students learning on real courses. Certainly student centred feedback should be used to determine speed of delivery. How many FE colleges are currently working on this with university research departments I wonder.

The Case for Digital Teaching Assistants

Georgia Tech has been experimenting with using Chatbots as teaching assistants and the story behind it is interesting. Named Gill Watson this was a teaching assistant that always seemed to be on the end of email and never tired of explaining things.

Initially Georgia Tech didn’t tell the students the teaching assistant was a bot!  Student reaction was fascinating as the next video shows.

The Chatbot Campus Genie

Not content with digital teaching assistants Deakin University has gone a step further with a Chatbot Campus Genie.

Similar to the Georgia Tech teaching assistant the technology used is based on IBM’s Watson. But this time the chatbot is designed to answer all those FAQs that myriad students ask year after year. The Genie is designed to give students a better experience and to save staff from the tedium of repeating the same answers time after time, year after year. Do happier students learn more effectively? Defining happy is only one challenge in answering this. But research by Daniel Gilbert at Harvard does indicate that happy employees earning more. Harvard based Research Schools International have showed a significant correlation between happiness and academic success in schools. So maybe the Genie will also impact academic outcomes to some extent.

I’d be interested to see research on this?

The Best AI Available?

The above are just a sample of what is currently available.  The details of these, and other teaching based software can be found with a few simple Google searches.

I’m sure that as technological advancement continues, better systems will be found and in six months time we’ll look back at the above as being the primitive first steps into digital learning. Change is likely to be rapid and far reaching. Rapid change however is not a reason to sit back and wait to see what develops. The more astute amongst us are already investigating and using technology similar to that I’ve mentioned. Those that are really on the ball are working with universities and making rapid progress for the institution and their students.

About Stefan Drew: FHE Marketing Consultant Stefan Drew was previously director of marketing at two FHE colleges and for the last decade has worked with colleges, universities and private providers throughout the UK, Europe and the US