Why teachers should help shape an age of abundance
Adam Sturdee | Starlight
Young people need adults who can look at an uncertain future and see something worth building. As teachers, we owe them an honest account of the risks around artificial intelligence. We also owe them the confidence that their choices can make a difference.
That second responsibility can get lost in the conversation. Job losses. Cyberattacks. Soaring energy demand. Existential risk. Add misinformation, deepfakes, privacy, bias, copyright and the concentration of power, and it is easy to feel that the future is a succession of threats heading towards us.
These concerns deserve serious attention. People whose livelihoods are disrupted need practical support. Children need protection. Powerful systems need scrutiny and accountability. A hopeful account of AI has to make room for all of that.
It also has to make room for what is becoming possible.
There is a reason the frightening stories are so hard to look away from. Human attention is sensitive to danger. The amygdala contributes to our responses to threat within a complex network of brain systems; the familiar description of a primitive “fear centre” oversimplifies the science. Our sensitivity to danger, though, gives alarming stories a powerful advantage.[1]
The newsroom saying “if it bleeds, it leads” has an online counterpart: if it worries us, we may click. A 2023 study in Nature Human Behaviour analysed around 105,000 headline variations tested by the publisher Upworthy. In those experiments, adding one negative word to a headline of typical length raised its click-through rate by 2.3%.[2]
That finding does not make every warning clickbait. Good journalism exposes harms that need exposing. It does explain an incentive: attention can be won by making us anxious. A headline can be accurate while the overall picture becomes distorted by the repeated selection of alarming stories. We should recognise when our concern is being used to keep us watching, clicking or scrolling.
Try changing the question. Search for “What positive news happened in AI today?” Then follow the links to the original research, check the dates and ask what has actually been demonstrated. Apply the same scrutiny to a cheerful announcement as to a frightening prediction. The point is to widen the evidence we see.
Consider a few examples.
On 16 October 2026, Penn Medicine reported research on an AI model that could identify signs of disease and future health risks in CT scans beyond the condition a scan was originally requested to investigate. The system is still being tested preclinically. Its potential is earlier recognition of illness from information already being collected.[3]
In March 2026, Google Research announced the rollout of AI forecasts for urban flash floods, reporting predictions up to 24 hours in advance. Better warning can give communities more time to prepare for a dangerous event.[4]
In biological research, a collaboration involving EMBL-EBI made millions of predicted protein complex structures openly available in 2026. These predictions help researchers investigate how proteins interact. They still need careful interpretation and experimental validation, but making them widely accessible expands the scientific work people can attempt.[5]
And there is evidence closer to the classroom. A 2024 randomised trial involving 259 teachers in 68 English secondary schools found that teachers using ChatGPT with guidance spent 31% less time preparing their Year 7 and 8 science lessons and resources. That meant about 25 minutes a week for the classes studied. Reviewers found no evidence of a difference in the quality of sampled resources. The trial examined preparation, rather than proving an improvement in pupil learning.[6]
These examples involve different kinds of AI and different stages of development. Together, they give us concrete reasons for hope: better tools for understanding disease, anticipating danger, investigating biology and supporting teachers. Each opens up possibilities for people to do something useful.
This is where the idea of an age of abundance becomes interesting.
In education, scarcity is familiar. There is never quite enough time to give every pupil another explanation, every colleague thoughtful feedback or every promising idea a proper trial. Access to specialist help depends too often on money, location or someone finding an extra hour.
AI offers a way to make some forms of assistance much more widely available. A teacher can develop alternative explanations to check and adapt. A small organisation can begin building a tool it previously lacked the skills to prototype. Someone facing a language barrier can get help understanding unfamiliar material. Useful support can become cheaper to provide and easier to reach.
The prospect is exciting because the gains could spread across so much of human activity. Better tools for research can support further discoveries; more accessible tools for creating things can bring more people into that work. The scale of that possibility deserves a place in our imagination alongside the risks.
Abundance, in this practical sense, means more people having the means to learn, create and solve problems. It could also mean more time for the parts of life that require our presence: a conversation with a struggling pupil, an unhurried coaching meeting, an evening with family.
Those benefits will depend on deliberate choices. Digital capability cannot remove physical limits or distribute prosperity by itself. AI has real energy demands, and the International Energy Agency warns that efficiency improvements sit alongside rapidly growing use and more demanding applications.[7] Access, affordability, ownership and the sharing of gains will all matter. If every minute saved simply becomes another task assigned, we will have missed a valuable opportunity.
Teachers have a part to play in those choices. We can ask whether a tool improves learning, whose needs it serves and whether it strengthens professional judgement. We can insist that time saved benefits pupils and staff. We can help shape school policies, challenge poor products and contribute evidence about what works.
We can also model an approach to uncertainty that young people can use themselves.
A pupil who encounters a problem can learn to ask: what would make this better, what can we test, and who could help? There is a useful entrepreneurial instinct in that question. It takes a difficulty seriously enough to look for a response. Teachers already encourage this whenever we help a learner move from “I can’t do this” towards a next step.
With AI, that might mean showing a class how to check a generated explanation against reliable sources and improve it. It might mean discussing when using a tool supports learning and when it avoids the thinking that learning requires. Strong subject knowledge gives pupils the basis for making those judgements.
For a teacher, a manageable beginning could be one school-approved task using non-sensitive material: drafting practice questions, then checking accuracy, usefulness and the time genuinely saved. Schools need to provide support and room to learn. Anxious or sceptical colleagues deserve to be heard; their questions can improve the decisions everyone makes.
At Starlight, we see this as a useful test of educational technology: does it help teachers understand their practice and give them more capacity to support learning? The value should be visible in the human work it makes possible.
There are good reasons to feel hopeful about the tools now within reach. We can explore ideas, make useful things and draw on forms of assistance that would have been difficult to imagine a generation ago. Young people deserve the chance to discover that possibility with adults who remain curious and exercise judgement.
Imagine a pupil leaving school with this conviction: there are difficult problems in the world, I can learn enough to understand them, and I can work with others to make things better.
That is a future worth teaching towards. An age of abundance is something we can help build, through the problems we choose to solve and the opportunities we choose to share. Our pupils should know that they have a place in that work.
Sources
Links supporting the evidence cited in the article. Checked on 29 October 2026.
- Fox and Shackman. An honest reckoning with the amygdala and mental illness. American Journal of Psychiatry, December 2024.
- Robertson and colleagues. Negativity drives online news consumption. Nature Human Behaviour, 16 March 2023.
- Penn Medicine. AI model can uncover unseen health issues in CT scans. 16 October 2026.
- Google Research. Protecting cities with AI-driven flash flood forecasting. 12 March 2026.
- EMBL-EBI. Millions of protein complexes added to AlphaFold Database shed light on how proteins interact. 16 March 2026, updated 19 May 2026.
- National Foundation for Educational Research. ChatGPT in lesson preparation — A Teacher Choices Trial. 12 December 2024.
- International Energy Agency. Key Questions on Energy and AI. Executive summary, 2026.
