From the fuss that is made about artificial intelligence in the media, you might imagine students to be getting an incredible online learning experience. But what is this so-called ‘AI revolution' looking like in schools?
It is almost the one year anniversary of an event that changed essay writing forever. When ChatGPT went public on November 30th, 2022, it was quickly followed by a torrent of speculation, wonder, and intrigue. Over a million people engaged with the language model within the first week of its launch, and ‘AI’ quickly became a buzzword for tech companies around the world, including those which ran online learning services. While students were busy getting ChatGPT to do their homework, companies were advocating their AI powered, intelligent learning systems to teachers. Nowadays I do my homework on several different platforms at a time. Some use AI, some don’t, while others could benefit from improvement.
Many students feel that online programs are easy to cheat, as there is nothing standing between an irritating question and a quick Google.
‘‘Most people just search it up’’ says Eshaal Mohsin, a year 10 student. “But it’s their responsibility to not cheat”.
Despite this, students are generally happy to use online programs for homework and revision, and teachers I have spoken to say that students are more likely to do homework if it is set online.
Talking of teachers, students are not the only ones using AI to their advantage.
“We get a lot more paperwork than we used to”, says Mrs A. Murrells, an English teacher. “And there are AI apps and stuff that can help deal with it.”
Nevertheless, many programs still have problems: bugs, people forgetting their passwords, and annoyingly picky questions. A particularly common complaint is the numerous test answers marked wrongly because of a typo, or a different way of phrasing something. AI could help, but even it struggles at complicated questions that require context.
One of my first experiences with ChatGPT was getting it to help with my maths homework. I thought I had finally found a calculator that could eloquently explain how to solve any problem. However, I soon realised that many of the answers were mathematically incorrect. When this was (politely) pointed out to the chatbot, it simply admitted that “You are correct, my previous response was not accurate.” This, even though ChatGPT is a language model, still shows how rudimentary our current AI systems are.
As it stands, we are at the very beginning of a long and treacherous climb up towards technological competence in education. While there are some concerns about AI replacing jobs in schools, it seems we are only just beginning to learn how useful it could be, for students and for teachers. We will need more people who are good at working discerningly and insightfully with AI. However, we are also realising that there will have to be plenty of changes before we get the full benefit of an AI education.
