The answer arrived in seconds. The understanding did not.
That is the question I want us to carry into our relationship with artificial intelligence. We are becoming remarkably good at getting a response. Are we becoming equally good at deciding whether that response deserves our trust?
I welcome a tool that helps a student untangle a difficult idea, a small organisation plan its work, or a writer find a clearer structure. Opportunity should not belong only to people who can afford a large team. But a smooth sentence can make weak reasoning look respectable. Convenience can quietly become authority.
UNESCO’s 2024 student framework places a human-centred mindset and ethics alongside technical understanding. Its emphasis on critical judgement is important. Knowing which button to press is a beginning; knowing what to question is a much deeper education. That is where I believe our schools, families and workplaces should put their energy.
Loading reader polls…
The American standards body NIST also treats generative AI as something requiring active risk management. A framework is not a guarantee that a particular answer is correct. It is a reminder to ask what can go wrong, who might be affected, and how we will notice before the damage grows.
For me, the practical test is straightforward. Can I explain the answer in my own words? Can I find the original evidence? Do I understand what remains uncertain? If I cannot, I have received an output, not acquired knowledge. That distinction matters whether we are choosing a course, preparing a proposal or making a public claim.
“Prove all things; hold fast that which is good,” says 1 Thessalonians 5:21 in the King James Version. I read that as a call to discernment. It does not ask us to reject every new thing. It asks us to examine what we receive, then keep what deserves keeping.
There is also a responsibility on the people building these tools. Users should be able to see limitations, question results and reach a human when a decision matters. A product that hides uncertainty behind confident language makes its user’s job harder, however attractive its interface may be.
I would like us to make room for the slower question in the meeting. The student who asks for evidence. The colleague who says the numbers do not fit. The reader who wants a source rather than another polished paragraph. These people are helping the work become stronger. They should not be treated as obstacles to speed.
My concern is not that a machine might be useful. It is that we might stop exercising the capacities that make its usefulness meaningful. Curiosity, patience, moral judgement and the courage to disagree still need practice. No subscription can do that practice on our behalf.
So use the tool. Explore what it makes possible. Then close the answer window for a moment and ask yourself what you actually think. The most valuable thing AI can help us produce may be a better question. We still have to be willing to ask it.
Evidence and reading
jai.a.sam
jai.a.sam is the founder of Students Herald and writes about international education, student opportunity and sector development. His columns explore the people, ideas and decisions shaping how students learn, travel and build their futures.
