Every now and then I find myself thinking back to what now seems a lifetime ago, namely 2020 and the pandemic. I find myself thinking about how this cataclysmic event affected our education and how we handled it.

On a personal note; in 2020 our family was mainly located in Sweden, Brazil, North Italy, New York and London – all places that made international headlines during the pandemic. At the start of 2020 I was dealing with severe sleep deprivation due to my then 1-year-old being awake 3-4 hours almost every night, and my wife being pregnant and having fainted once on the metro I took all the difficult nights. By February I was exhausted enough that I asked my boss if I could take a few days off to rest – but swiftly had to interrupt my vacation to go in and write a plan for continued teaching during a pandemic for the school.

We wrote, discussed and tried. We set up a self-hosted solution with BigBlueButton as well as tried Google’s and Microsoft’s solutions (at the time, BBB easily won out as the one best suited for educational purposes, but we kept the others as backup solutions). We tested with a few students, then a full class during the day and the, to the horror of people higher up, we decided to run a full-scale test day for distance teaching. We reasoned that we needed a test run for both teachers and students just in case we would need to transition to distance education, so we tested and the next day evaluated among teachers and students, as well as made much needed improvements to the technical solution.

The day after that, we were ordered along with the rest of the upper secondary schools (gymnasieskolor) in Sweden to go distance. A few weeks after the move to distance education we had documents providing technical support to teachers as well as educational tips to both teachers and students (the first drafts of these were done before the order came, but required some iterations before they could be considered done). A few months after, every single classroom in the school was outfitted with a ceiling-mounted camera, speaker and microphone and a desktop setup which made all the technology required for distance or hybrid teaching plug-and-play.

I’m not saying it was easy, or without conflict, but considering that education is often seen as almost immovable – often (at least in Sweden) described as ”turning an oil tanker” – it is worth remembering that in 2020 education turned around swiftly and considering the challenge relatively smoothly. We had discussions and worries, and there were definite problems – but we did it. I know my view is slightly skewed by having worked at a school where we had highly skilled people that could support the technological aspects of the transition, but even in a broader outlook, a remarkable transition was made.

Which brings me to today. AI may not be the immediate cataclysmic event that a pandemic is, but its effect on education has been profound during 2023. I cannot help but reflect on the difference in how education has responded in these two situations. In 2020 schools could swiftly set up or gain access to the technology required to meet the needs of distance education – we already had learning platforms, and most schools already had access to Google Meet or Microsoft Teams. Today, most schools I have any contact with still haven’t got any official access to AI, with frustration and worry growing among teachers.

In 2020, we could fairly quickly make recommendations, guides and support teachers to get them up and running in this new situation, but now most schools are still lacking any guidance within or from above. This is not to say that there haven’t been swift and clear support from other organisations, like UNESCO, about AI in education, but very little has happened on the local level.

One lesson I took from the experiences of 2020-2021 is that education doesn’t have to be slow to respond to new challenges. One major difference, I think, between now and then is that during the pandemic the need was immediate and we were given the freedom to solve the problems locally, while today many schools are waiting for direction. It’s partly the economy – there is no money to actually give the people with knowledge any time to actually help out. It’s partly that this is more of an unknown, and there is a lot of worry and insecurities both when it comes to the technology and legal questions. We could, honestly, have rather quickly set up an interface for AI models that teachers could start using first for their own development and then for trying in the classroom. We could have provided support, guides, tips directions for schools many months ago, with the caveat (which I think everyone would accept) that they would be a work in progress and subject to change and updates – but which could still act as support for getting started.

I’ve been talking to, held lectures for and workshops with teachers, and there is a widespread feeling that we have to work with this, but teachers need access, time and support. And that is one major reason why education is slow to change – because change requires effort, which requires support and time to develop, which is more or less nonexistent in this case.

Another aspect where a comparison may be valid is that with the pandemic there was at least initially a sense of teachers and students (and other staff) working together, finding their way together in collaboration against the situation. Later came discussions on risks of cheating and how to measure knowledge fairly but securely.

With AI (or specifically ChatGPT since that was the catalyst) the initial general reaction was almost the complete opposite. First came suspicion, risks of cheating and of losing control of what students do.

2020 started in collaboration and coming together, while 2023 to a larger extent drove us apart. This too makes the transition and development more difficult.

Other lessons from teaching during the pandemic

Here are a few more thoughts from the pandemic:

Throughout the pandemic and after, we asked students how they fared. While we asked and analysed data in an organised manner, it was not collected for research (unfortunately – now in hindsight I wish I had thought of collecting this data for research purposes in addition to ongoing evaluation and development). As can be expected, there was a group who struggled with the transition to distance and later hybrid education and there was a large group who felt that they did about as well. What I find interesting and lately frustrating is the group that was steadily around 10-15% that claimed to fare better. This was especially true during the longer period of hybrid education that we had after the first shutdown. At the school I worked at then, we tried and evaluated different models and finally decided on splitting each class in half, and alternating which half was at school and which half attended virtually. While some teachers found this quite a bad experience (however it was agreed upon as the best alternative available), it suited me quite well. I found that I was a better teacher with students in the classroom than fully virtual, but the division had several advantages.

For example, we noticed quickly going to virtual that students that tended to disappear in the classroom suddenly started asking and answering questions in the chat. Students that before was in the background sometimes came forward in the virtual classroom and answered questions from other students more quickly than we teachers – a pattern that at least to some degree held true also in the hybrid classroom. Students reported being better able to focus both in the classroom and at home during the hybrid mode of education, citing smaller groups and alternating between virtual and classroom as advantageous to focus and providing a better balance in their needs for self-reflections and focus, and the necessary social aspects of school that comes with being on-site. This said, there was definitely a group of students for which even the hybrid mode was clearly detrimental, and who did not manage to learn at all at home.

I do think, now, about that group of students who clearly and consistently reported doing better during the pandemic, and why we completely abandoned that group of students once we could go back to business as usual. Both during the pandemic and after, and not only in Sweden, I talk to students who wish for elements of the pandemic education experience to return, maybe a day or two a week. To be fair, the same may not be true for teachers – but I still wonder if we shouldn’t at least have a discussion of whether there is room for some aspect of hybrid teaching in our system. I think for some students, it may improve both their studies and health.

This may be even more true with the advancements in AI, where the improvement that some students experienced in virtual teaching may be augmented by an AI teacher, which may also balance the added cognitive load that some teachers experienced in the hybrid mode.