Pick a prompt, switch between band levels, and see exactly how a Band 8+ response differs from a Band 6 or 7. Curated from the current 2026 exam pool with higher-band vocabulary highlighted for reference.
Some people believe that technology has made students lazy and less able to think for themselves. To what extent do you agree or disagree?
In many cities, private cars are being restricted in favour of public transport. Discuss the advantages and disadvantages of this policy.
Generative AI tools are increasingly used to complete workplace tasks that were previously done by humans. Why is this happening, and does it have a positive or negative effect on the workforce?
Many people say technology makes students lazy. I agree with this idea because students use a lot of tools now. For example, they use calculators and the internet for homework. This means they don't think a lot. However, technology also helps students learn new things. In my country, many kids watch educational videos. So, technology is both good and bad for students.
It is often argued that modern technology has reduced students' willingness to think independently. I partly agree with this view, as over-reliance on digital tools can weaken critical thinking, though technology itself is not the root cause. On the one hand, students who use search engines or AI assistants may accept the first result they find without questioning it. On the other hand, when used carefully, technology gives learners access to a vast range of sources and encourages deeper analysis. In conclusion, the problem lies in how technology is used rather than in the tools themselves.
It is frequently claimed that the pervasive use of digital tools has eroded students' capacity for independent thought. I largely agree with this proposition, although I would argue that the issue reflects pedagogical habits more than technology itself. To begin with, an over-reliance on search engines and AI assistants can undermine analytical reasoning: learners now expect instant, packaged answers rather than engaging in sustained reflection. Nevertheless, technology also facilitates deeper inquiry when it is embedded within a rigorous curriculum. In conclusion, whether technology fosters or hinders thinking depends less on the tool and more on how educators frame its use.
The contention that digital technology has dulled students' intellectual autonomy is one I find largely persuasive, though the causal picture is more nuanced than critics allow. Undeniably, the frictionless availability of search engines and generative AI has cultivated a habit of intellectual outsourcing: rather than wrestling with a problem, learners increasingly retrieve pre-formed conclusions. Yet to lay the blame at technology's door is to overlook the curricular and cultural context in which it is deployed. Where teachers scaffold enquiry — insisting on source triangulation, transparent reasoning and iterative revision — the same tools become catalysts for deeper thinking rather than substitutes for it. The pathology, in short, lies not in the technology but in an education system that has not yet learned to teach alongside it.
Nowadays many cities want less cars. This has good and bad points. The good thing is less pollution and less traffic. Buses and trains can carry many people. The bad thing is that some people need cars for work or family. Also public transport is not good in every city. In my opinion, cities should improve buses first and then restrict cars.
In recent years, a growing number of cities have introduced measures to limit private car use and expand public transport networks. While this policy offers clear environmental and social benefits, it also raises concerns for certain groups. On the positive side, restricting cars reduces congestion and air pollution, making city centres more liveable. On the negative side, families in suburbs and workers with irregular schedules may struggle without their vehicles. Overall, the advantages outweigh the drawbacks provided that reliable alternatives are available.
A growing number of municipal authorities are actively curbing private vehicle use in favour of expanded public transport networks. While such policies deliver substantial environmental and civic gains, they also impose transitional costs that policymakers cannot afford to ignore. The principal advantages are well documented: lower emissions, less congestion and more walkable urban cores that in turn support local commerce. Conversely, sudden restrictions can disproportionately burden low-income commuters whose neighbourhoods remain poorly served by transit. On balance, the policy is defensible, but only where investment in reliable, affordable alternatives precedes any restriction of private cars.
Around the world, city governments are quietly rewriting the terms on which the private car is permitted to dominate urban space. That this trend brings substantial civic dividends is beyond serious dispute; the more interesting question is how equitably those dividends are distributed. The upside is straightforward and well evidenced: cleaner air, quieter streets, safer pedestrians and a rediscovered appetite for street-level commerce. Yet the same policies, applied bluntly, can entrench existing inequalities — punishing those whose peripheral neighbourhoods have long been underserved by transit and who therefore have no realistic alternative to driving. A defensible restriction regime, then, is one built on sequencing: first the buses, the bike lanes and the reliable timetables; only then the congestion charge.
Nowadays many companies use AI tools like ChatGPT for work. This happens because AI is fast and cheap. Bosses want to save money and time. In my opinion, this is both good and bad. It is good because workers can finish jobs faster. It is bad because some people can lose their jobs. So companies should train workers to use AI, not replace them.
In recent years, generative AI has been widely adopted at work because it dramatically cuts the time and cost of routine tasks such as drafting emails, summarising reports and writing code. Employers naturally want the productivity gains that follow. The impact on the workforce, however, is mixed. On the positive side, employees can offload repetitive work and focus on higher-value thinking. On the negative side, entry-level roles in fields like copywriting and customer support are already shrinking. Overall, I believe the effect is positive only where companies retrain staff rather than simply cutting them.
The rapid uptake of generative AI at work is driven primarily by economics: models such as GPT-class assistants can perform, in seconds, drafting and analytical tasks that once absorbed hours of skilled labour. Firms facing tight margins have every incentive to adopt them. The net effect on the workforce, however, is far from uniform. For senior professionals, AI acts as a force multiplier, offloading routine cognitive work and freeing capacity for judgement-heavy tasks. For junior and administrative staff, the picture is bleaker: the very tasks that once served as career on-ramps are now the first to be automated. On balance, the outcome is positive only where organisations pair adoption with genuine reskilling; without it, gains in productivity will come at the cost of narrower entry routes into skilled work.
Generative AI has moved into the workplace at a pace few technologies have matched, and the reasons are almost entirely economic: a general-purpose model can now discharge, in seconds and at negligible marginal cost, the drafting, summarising and coding tasks that once justified whole teams. Whether this constitutes progress depends less on the technology than on how firms choose to metabolise it. Used as a complement, AI amplifies experienced judgement and liberates senior staff from clerical drag; used as a substitute, it quietly erodes the junior roles through which that judgement was traditionally acquired. The genuine risk, then, is not mass unemployment but a hollowing-out of the professional ladder — a workforce in which there are seats at the top and at the entry point of the algorithm, but progressively fewer rungs in between.