1.1 Potentials & Risks
Take Home message
Should AI come with a leaflet like a medicine?
You know how every medicine comes with a little paper? The one you probably just throw away? It's a pretty boring document, and it's one of the under-the-radar achievements of the last century. It tells you four things: what the medicine is for, how much you need to take, what it will do to you that you didn't ask for and who shouldn't take it at all.
There's always an apparatus behind these lines: trials, a regulator and a duty to publish any missteps. No one thinks this leaflet makes the medicine work. It basically means the medicine can be used by anyone, even if they weren't involved in its development. That's the whole point, and it's exactly what we're missing when we are using AI as teachers.
The AI tools now quietly occupying your school's devices showed up without so much as an instruction sheet. In the UK a review found, that only 7 out of 100 education technology companies had ever tested their own products in a controlled trial, and just 12 out of 100 had sought external certification. The story is no different in the classroom. Only 11 out of 100 school administrators and teachers across 17 American states said they'd ever asked for peer-reviewed evidence before bringing in a new tool into school. Since no one asked, no one had to explain.
And the outcome is as plain as day: wasted money. In the US, two thirds of school software licenses sat idle, and 98% were barely touched. The EdTech Genome Project examined around 7,000 teaching tools worth 13 billion dollars and found that 85% were either mismatched or misused. Not scams. Not defective. Just tossed into the hands of people left without a leaflet.
This chapter might seem like the plain, obligatory part, but it uncovers the real stories behind the playbook: It reports what truly happened, who felt the impact, and points to the side effects, even those that quietly touched children who never volunteered.
What is the AI medicine for?