The phrase “future gazing” has came up recently so I thought it worth sharing some thoughts on the future of EdTech as I see them. As such I intend to share a series of separate posts on different technologies which might have an impact on education in the years ahead.
This is a big topic in the wider IT world but also increasingly in education. The challenge is that AI covers a multitude of sins plus the application of the different AIs are substantial.
The holy grail of AI, as I see it, is the general-purpose AI. Am not going to spend any real time in this area as this is, in my opinion, some way off. When it does become a realisation, there is great potent
ial for it to be used in education to supplement teaching staff both as a virtual teacher, a virtual classroom assistant or a virtual coach or mentor. As I said however, this is some years off.
More specific purpose AIs are much more likely to make an appearance in the short term. An example of this might be a Mathematics AI which students can pose questions to in natural language, and that will then either provide answers or direct students to appropriate learning materials. This isn’t that far off and is being used already on organisations help pages. It just hasn’t thus far been focused on education.
Another application of AI might be in its ability to recognise emotions and activities of students. This is already in trial in China. Basically, this involves a classroom camera and an AI which analyses the facial expressions of students along with what they are doing. This information is then fed back to the teacher to inform learning. The teacher will get information on the students which appear confused or upset, indicating possibly they are struggling with the materials, along with data on which pupils have been busy with the work, which have been raising their hands to ask questions or provide answers and those which have been more disengaged or not participating. From this the teacher can then decide how to change the learning activities, target questions or revisit concepts. I suspect this AI could also be expanded to look at teacher questioning and provide feedback and advice on the types of questions being asked, the frequency and who the questions are directed to. It might also look at the engagement of students throughout the school day to try and identify trends and develop a structure for the school day which is more in line with the physiological and psychological needs of students.
School data analysis is one area where I think AI is very close to being usable widely in schools. Schools already are sat on a wealth of data in terms of the student academic data, student demographic data and pastoral data among others. AI or machine learning can easily analyse the data and identify patterns which humans may not be able to identify. At a school level this can easily be applied to summative academic results, identifying how different student groups perform, allowing comparisons across subjects, etc, however as we gather more and more formative data these AIs will then be able to feedback to teachers in relation to areas which students do or do not understand. It will also be able to identify whether there is a pattern across different teachers therefore suggesting a change to how a particular topic is taught, or whether it relates to a group of students or to specific related topics.
In the wider school there will also be opportunities for use of AI. In the dining hall for example AI might be able to examine data to identify possible lunch timings to improve efficiency. Analysis of book titles taken from the Library might help in providing a window into pupil preferences and interests. AI may have the ability to examine parents evenings and parents meetings to try and streamline these events and ensure everyone gets to see who they need to see with a minimal period of waiting. Machine learning may be able to examine teacher performance management data and identify opportunities for peer support and peer learning to occur, or to identify cross school professional development needs. Facilities use might be analysed to identify when they are under-utilised and then seek to make them available to the local community. Teacher work days might be optimised through AI recommendations resulting from an analysis of our working habits looking at when we tend to send emails, our timetable, who we commonly meet with, etc. These are just some of the ways in which AI may makes its way into our school.
Artificial Intelligence is going to make an increasing appearance in schools. I think this is inevitable. In actual fact I would say to some extent AI or Machine Learning is already in schools possibly in the schools firewall or mail filtering solutions or in the network infrastructure. Going forward however it will become much more visible as it enters more areas of school life. Or maybe like all good technology use, may become more common yet will be transparent in its use, with users unaware of where AI is providing help, support and guidance.
I think the general-purpose AI, the Data from Star Trek TNG, or HAL for 2001: A Space Odyssey is some way off. In the first instance AI will provide hints and tips as well as other low-level recommendations or suggestions. It is to this, and the possible productivity and efficiency gains that may result, that we should therefore first look.