IELTS Academic Reading — Practice Test 7

Few technologies have been promised with as much confidence as the self-driving car. Its champions argue that machines, unlike humans, do not grow tired, drive drunk or glance at a phone at the wrong moment, and since the great majority of road deaths are caused by human error, removing the driver should in principle save many lives. The engineering has advanced rapidly, and limited fleets of autonomous taxis now operate on the streets of several cities. Yet as the vehicles move from the laboratory to the road, it has become clear that the hardest questions they raise are not about sensors or software but about ethics, responsibility and the shape of society.

The dilemma most often discussed is also, arguably, the least realistic. Philosophers have long posed the 'trolley problem', in which a runaway tram will kill several people unless it is diverted onto a track where it will kill one. Transferred to the road, the puzzle asks how a car should behave when a collision is unavoidable: should it protect its passengers at all costs, or sacrifice them to spare a larger number of pedestrians? Critics reasonably observe that such clean, either-or situations almost never arise, and that a well-designed car should brake long before matters reach that point. Even so, the thought experiment exposes something real. A human driver reacts in an instant, without deliberation; a programmed car embodies choices made months earlier by engineers, and those choices encode a set of values that someone must decide upon and defend.

Closely related is the question of blame. When a conventional car crashes, the law knows where to look: the driver was careless, or the manufacturer sold a faulty part. An autonomous vehicle blurs these lines. If the software misjudges a situation, is the fault that of the company that wrote the code, the owner who failed to update it, or the regulator who approved it? Manufacturers have an obvious incentive to frame any accident as misuse by the owner, while owners will point to the machine. Some legal scholars argue that traditional notions of fault may have to give way to a system in which manufacturers accept liability by default, treating collisions as a cost of doing business much as airlines treat rare mechanical failures.

There is also the matter of what these vehicles know. To navigate, an autonomous car must sense its surroundings continuously and remember where it has been, generating a detailed record of every journey its user makes. That information is valuable — to advertisers, to insurers, and to the state. A car that logs each destination is, in effect, a tracking device on wheels, and the temptation to use its data for purposes far removed from transport will be considerable. Whether such records are protected as private, or traded and demanded like any other commercial asset, is a decision that current law is poorly equipped to make.

Beyond the individual vehicle lie broader social effects, and the most immediate concerns work. Driving is among the most common occupations in many countries; millions earn their living at the wheel of a lorry, taxi or delivery van. If the technology matures, a large share of those jobs could disappear within a generation. Optimists note that earlier waves of automation eventually created new kinds of employment, and that the transition, if it comes, will be gradual rather than sudden. That is little comfort to a middle-aged driver whose skills are made redundant, and the burden of such change tends to fall hardest on those least able to absorb it.

The promise of autonomous vehicles is often framed in terms of freedom and access — mobility for the elderly, the disabled and those who cannot drive. Whether that promise is kept depends on how the technology is deployed. If self-driving cars are cheap and shared, they might reduce the number of vehicles on the road and widen access to transport. If they are expensive and privately owned, they could do the opposite, and by making travel effortless they may even encourage people to take more and longer trips, worsening the congestion and emissions they were meant to relieve. The same machine, in other words, could make cities more equitable or less so, depending on choices that have nothing to do with the vehicle itself.

It is tempting to treat the remaining obstacles as merely technical — a matter of teaching the car to handle snow, or to interpret the gesture of a police officer waving traffic on. But the deepest challenges are of a different order. They concern how much risk a society will tolerate, how it assigns responsibility when things go wrong, and who gains and who loses as the technology spreads. These are not problems that better sensors can solve. They are questions about values, and they will be answered, well or badly, by the public and its representatives rather than by engineers alone.

Most of us think of memory as a kind of recording. We imagine that experiences are captured and filed away, and that to remember is to retrieve a stored copy, faithful if sometimes faded. This intuition is deeply held and almost entirely wrong. Over the past century, research has steadily replaced the image of memory as a video archive with something far stranger: a reconstructive process in which the brain assembles a plausible version of the past each time it is called upon, drawing on fragments, expectations and general knowledge as much as on any fixed trace. Remembering, on this account, is less like replaying a film than like rebuilding a scene from a handful of clues — and, crucially, the rebuilding can go wrong in ways the rememberer never notices.

The first systematic challenge to the recording model came from the British psychologist Frederic Bartlett in the 1930s. He asked English volunteers to read an unfamiliar Native American folk tale, 'The War of the Ghosts', and to retell it after various intervals. Their reproductions were revealing. Details that made no sense within the readers' own culture were quietly dropped or altered; supernatural elements became rational, and the story grew shorter, tidier and more conventional with each retelling. Bartlett concluded that people do not store experiences intact but fit them into pre-existing mental frameworks, which he called schemas. Memory, he argued, is governed by 'effort after meaning' — a drive to make the past coherent, even at the cost of accuracy.

If Bartlett showed that memories drift over time, later researchers demonstrated how easily they can be pushed. In a series of influential studies from the 1970s onward, the American psychologist Elizabeth Loftus showed people film of a car accident and then questioned them about it. When she asked how fast the cars were going when they 'smashed into' each other, witnesses gave higher speed estimates — and were more likely to report broken glass that had never appeared in the film — than those asked how fast the cars were going when they 'hit'. A single word, planted after the event, had altered what people believed they had seen. This 'misinformation effect' has since been reproduced many times, and it carries an uncomfortable lesson: the very act of questioning a witness can reshape the memory it is meant to recover.

More disquieting still is the discovery that whole events can be invented. In a famous demonstration, Loftus and her colleagues persuaded a substantial proportion of participants that, as children, they had once been lost in a shopping mall — an episode that had never occurred. Given a gentle suggestion from a trusted relative and asked to describe the event repeatedly, many people not only came to believe it but supplied vivid, confident details of their own. Later studies implanted richer and more improbable memories still. The lesson is not that everyone is equally suggestible, for they are not, but that the line between a genuine recollection and a manufactured one is far blurrier than common sense allows, and that confidence is no guide to which is which.

Why should a faculty so central to our lives be built to err? One answer is that we have misunderstood what memory is for. On an increasingly influential view, its function is not to preserve the past for its own sake but to help us act in the future, and a system optimised for flexibility will inevitably sacrifice fidelity. Recalling the gist of an experience — what it meant, what it implies — is often more useful than a verbatim record, and the same machinery that lets us generalise, imagine and plan is the machinery that fills gaps with plausible invention. On this reading, false memories are not a malfunction but the price of an otherwise remarkable capacity. Not everyone is persuaded. Some argue that this functional story, however elegant, risks excusing the failures rather than explaining them, and that the reliability of ordinary, everyday memory should not be understated simply because it can be manipulated under laboratory conditions.

The stakes of this research are highest where memory meets the law. For most of the twentieth century, the confident testimony of an eyewitness was regarded as close to decisive in a courtroom, and juries found it hard to disbelieve a witness who was sure. The study of memory has undermined that confidence. Analyses of wrongful convictions later overturned by other evidence have found that mistaken identification by sincere, honest witnesses was among the leading causes. The same findings have fuelled a bitter dispute over 'recovered' memories of childhood abuse, allegedly retrieved in adulthood during therapy: while some such memories may be authentic, the mechanisms that can implant false ones make any recollection produced by suggestive questioning deeply suspect.

There is something unsettling in all of this, for we tend to treat our memories as the bedrock of who we are. If the past we carry within us is partly a construction, revised each time we revisit it and vulnerable to the suggestions of others, then the self built upon it rests on less solid ground than we like to think. Yet the same flexibility that makes memory fallible is inseparable from what makes it powerful: the ability to draw lessons from experience, to picture what has never happened, and to imagine a future different from the past. A perfect recording would, in a sense, be less useful than the imperfect, creative faculty we actually possess. The task, then, is not to lament that memory is reconstructive but to understand its limits — and to be appropriately humble about the certainty of our own recollections.