IELTS Academic Reading Practice Test 132

In the mid-1950s the two superpowers were racing to build rockets that could cross oceans. The Soviet R-7, a missile designed to carry a nuclear warhead, was powered by a cluster of liquid-fuelled engines and was first tested successfully in August 1957. Both the United States and the Soviet Union had also promised, during the International Geophysical Year of 1957–58, to put a scientific satellite into orbit. Neither programme was driven by curiosity alone. For Sergei Korolev, the chief designer of the Soviet rocket, the research was genuine, but the political prize was greater still.

Korolev's team had spent several years on an ambitious satellite known as Object D, a laboratory the size of a small car and packed with instruments to measure cosmic rays, temperature and the Earth's magnetic field. By the summer of 1957 it was clear that the finished spacecraft would be far too heavy for the rocket that was ready to fly, and that many more months of work would be needed. Fearing that the Americans might launch first, Korolev obtained permission to send up something much simpler, a device that would demonstrate only that a satellite could be placed in orbit at all. Object D was eventually launched as Sputnik 3 in May 1958, by which time it was no longer a first.

The substitute was a polished sphere of aluminium alloy 58 centimetres in diameter, weighing 83.6 kilograms. Its sealed interior was filled with nitrogen at a pressure slightly above that of the atmosphere, so that any puncture by a micrometeorite would be recorded by instruments. Two radio transmitters, powered by batteries, sent out a simple pulsed signal. Four antennas, two of them almost three metres long, were fixed to the outside of the sphere so that the signal could be broadcast in every direction. The design was deliberately plain: the satellite carried no camera and almost no scientific equipment.

Sputnik 1 left the launch site at Baikonur, in what is now Kazakhstan, on 4 October 1957. The rocket placed it in an elliptical orbit between roughly 215 and 940 kilometres above the Earth, which the satellite completed every 96 minutes. The world heard about it almost at once: the transmitters broadcast on frequencies of 20.005 and 40.002 megahertz, and radio amateurs who tuned in could listen to the beeps for themselves. No announcement had been made beforehand. Observers with binoculars could follow the faint moving point it made as it crossed the sky near sunrise and sunset.

Orbiting through the thin upper atmosphere, the satellite gradually lost height. Its batteries failed after three weeks, and the beeping stopped. By then it had circled the Earth about 1,440 times and travelled some 70 million kilometres. On 4 January 1958 it fell into the denser air, heated up and burned away.

The reaction in the United States was close to panic. American intelligence had expected the Soviet programme to be slow, and few officials had believed that a satellite would appear so soon; the shock was sharpened by the fact that the R-7 could plainly deliver a warhead to any city in the world. The government created the Advanced Research Projects Agency, spending on the teaching of science and mathematics was increased, and in July 1958 the National Aeronautics and Space Act established NASA, which began operations that October. The episode became known as the Sputnik crisis, and it changed the terms of the competition.

A second satellite followed on 3 November 1957. Sputnik 2 weighed 508 kilograms and carried a dog, Laika, the first animal to orbit the Earth. The spacecraft had been assembled in a few weeks and had no means of returning, so the dog was always going to die in space; accounts published many years later showed that she had survived only a few hours before overheating killed her. The flight provoked argument about the treatment of animals in research, and the satellite itself broke up on re-entry in April 1958. The Soviet Union released few technical details at the time.

The American answer came on 31 January 1958, when Explorer 1 was launched by a team that included the engineer Wernher von Braun, with instruments designed under the physicist James Van Allen. A Geiger counter on board showed that the spacecraft was passing through bands of intense radiation trapped by the Earth's magnetic field, a discovery that became known as the Van Allen belts. A much smaller satellite, Vanguard 1, followed on 17 March 1958 and was the first to be powered by solar cells; it remains in orbit today. Explorer 1 weighed only about 14 kilograms, yet its instruments produced a major scientific result.

Within a few years the technology had changed from a spectacular demonstration into ordinary infrastructure. Thousands of active satellites now circle the Earth, alongside many that have been retired, and they are used for communication, navigation, weather forecasting and scientific research. By measuring the drag on early satellites, scientists also learned how dense the upper atmosphere is, a region too thin for balloons to reach and too high for aircraft. Most of us use the results of the space age every day without a thought for the small sphere whose beeps began it. Its launch is now treated by historians as the opening of the rivalry between the two superpowers in space.

Wearable health devices have moved in two decades from research laboratories to the wrists of hundreds of millions of people. The commonest form is a watch or a ring containing light-emitting diodes, a light detector, an accelerometer and a small processor with a radio link to a phone. The same electronics that count steps can estimate heart rate, breathing rate, the oxygen content of the blood and the pattern of sleep. Chest straps and adhesive patches record the heart more accurately, and the boundary between a fitness gadget and a medical instrument is increasingly hard to draw.

Most wrist devices measure the pulse optically, using a technique called photoplethysmography. Light-emitting diodes shine light into the skin, and a photodiode beside them measures how much of it comes back. Green light is used for the pulse because haemoglobin, the pigment that carries oxygen in red blood cells, absorbs it strongly. Each time the heart contracts, arteries in the wrist expand and the volume of blood under the sensor rises, so less light returns. Between beats the volume falls again and more light returns. The result is a wave with a peak for every heartbeat, from which the pulse rate can be counted and the rhythm examined. The same signal contains slower variations caused by breathing, which is why many devices also report a respiration rate.

Blood oxygen is estimated in a similar way, but with two colours of light. Oxygen-rich and oxygen-poor haemoglobin absorb red and infrared light differently, and the ratio between the two signals gives an estimate of oxygen saturation, the percentage of haemoglobin that is carrying oxygen. Clinical pulse oximeters, clipped to a finger, have used this principle since the 1970s. Shrinking the measurement into a watch was a considerable engineering achievement, and it depends on firm contact between the sensor and the skin.

A different approach records the heart's electrical activity. Each contraction begins with an electrical impulse, and a watch can pick up the tiny voltages that this produces on the skin. In one common design the wearer rests a finger on the watch's crown, completing a circuit across the body. The resulting trace is a single lead, much less detailed than the twelve-lead electrocardiogram recorded in a hospital, but enough to show whether the heart is beating in a regular rhythm. In 2018 the American regulator cleared one watch function for detecting atrial fibrillation, an irregular rhythm that raises the risk of stroke. A study of nearly 420,000 users, published in 2019, found that about 0.5 per cent received such an alert.

Beyond fitness, the devices are being used in medical care. Patients with heart failure, who may need frequent checks of weight, pulse and breathing, can be monitored at home, and an abnormal reading can prompt a call from a nurse before symptoms become severe. Adhesive patches worn for a few days are used to look for irregular rhythms that a short appointment would miss. In clinical trials, wrist devices can supply continuous measurements of activity and sleep that participants would struggle to record accurately themselves.

None of this is as precise as it looks. Optical readings are disturbed by movement, so a watch may report a pulse of 150 during vigorous exercise when the true rate is far lower, or lose the signal altogether. Melanin, the pigment that darkens skin, absorbs green light, and tattoos and scars can block it, so readings may be weaker for some wearers than for others. Several studies have found that consumer sleep trackers agree only moderately with polysomnography, the laboratory method that records brain waves, eye movements and muscle activity, and that they tend to overestimate total sleep time.

Regulators have begun to draw distinctions. A product that claims only to encourage exercise is treated as a general wellness item and needs no approval, whereas one that claims to detect a disease is a medical device and must be shown to work. In 2021 the American authorities warned that pulse oximeters, including those used in hospitals, might be less accurate in patients with darker skin, a problem that had been reported for years. The warning was a reminder that a sensor tested on a small sample may behave differently across a whole population.

Research is now directed at measurements that still require a needle or a cuff. Blood pressure is the most sought after: instead of a cuff, a device can time the delay between the electrical impulse that starts a heartbeat and the arrival of the pulse in the wrist, since that delay shortens as pressure rises. The method is promising, but it has not yet been shown to be reliable enough to replace a cuff. Other groups are developing sensors that analyse sweat for glucose, lactate and sodium, and flexible materials that can be worn on the skin like a plaster. Non-invasive glucose measurement remains the hardest problem of all: several well-publicised attempts have failed to convince independent researchers.

Wearable devices are best understood as instruments that extend the reach of medicine into everyday life rather than as replacements for it. They can reveal a trend over months that a single appointment cannot, and they can prompt someone to seek help. But a number on a screen is not a diagnosis, and the most useful thing a clinician can do with it is to ask what it means for this particular patient.

Every few months a new program produces something that would once have been called art: a portrait in the manner of a Dutch master, a passable sonnet, a piece of music that moves listeners. Each time, the same argument breaks out. One side insists that the machine is merely rearranging what it has been given; the other replies that human artists do much the same. The disagreement is seldom resolved, because the question at its centre is usually posed in a way that makes an answer impossible.

If the question is to be answered at all, the word itself needs unpacking. A creative work is normally expected to be both new and valuable: a random arrangement of words is original but means nothing, while a competent imitation of a familiar style is valuable but adds nothing. The cognitive scientist Margaret Boden has distinguished three kinds of creativity, a scheme more useful than any single yes-or-no test. Combinatorial creativity puts familiar ideas together in an unfamiliar way. Exploratory creativity works within an established set of rules and finds possibilities that their originators had not noticed. Transformational creativity alters the rules themselves, so that what counts as good in a field changes.

Combinatorial creativity, in Boden's scheme, is within reach of any system that can search a large space of combinations, and machines have become extremely good at exactly that. The most discussed example came in 2016, when a program called AlphaGo played the second game of a match against the professional Go player Lee Sedol. On the thirty-seventh move, AlphaGo placed a stone where the experts watching did not expect it, and some commentators assumed at first that it had blundered. The move proved important, AlphaGo won the game, and Lee Sedol later described the program's play as creative.

The sceptical case is older than the technology. In the 1840s Ada Lovelace, writing about Charles Babbage's planned Analytical Engine, observed that a machine of this kind could only do what it was ordered to perform and could not originate anything. Modern systems do not obviously refute her. A language model does not build a picture of the world and then describe it; it learns statistical patterns from an enormous body of text and produces whatever is most likely to follow. When a generated poem moves a reader, the feeling is real, but the intention behind the poem is not.

Yet the romantic picture of the solitary genius, struck by inspiration and answerable to nobody, is a recent invention. Most creativity is social. Painters work within traditions and depend on dealers and critics; scientists build on the findings of others and are judged by their peers. What counts as creative is decided by communities that recognise, interpret and argue about the work. That is why authorship becomes so tangled when a machine is involved. In 2018 a portrait produced by a small collective using a generative network was sold at auction for more than 400,000 dollars, and the discussion that followed concerned not the image but the names attached to it.

We are also inconsistent in how we judge. A human work arrives with a history — the years of practice, the ambition, the failures in other pieces — and we see it through that history. A machine's output arrives with nothing but the prompt that produced it, and is inspected for its flaws. This asymmetry is not unreasonable, since the history is part of what we value, but it means that comparisons between the two are rarely fair tests of the machines. It suggests something about creativity itself: that creativity is not simply a property of an object, but something we grant to agents whose purposes we think we understand.

The practical consequences are arriving faster than the argument. Images, translations, routine journalism and design drafts can now be produced in unlimited quantities at almost no cost per item, and illustrators and photographers report that clients ask them to compete with prices no person can match. Copyright law, which was designed around identifiable authors and original works, has struggled to say who owns a machine's output, and legislators have moved slowly. There is a further risk of homogenisation: a system trained on the past will reproduce the conventions of the past, and if much new work is made that way, the range of what is available may narrow.

My own view is that the argument about whether machines are creative is mostly beside the point. These systems already produce results that would be called creative if a person had produced them, and there is no reason to deny that something like combinatorial and exploratory creativity is present in what they do. But they have no stake in what they make. They do not belong to a tradition, cannot be disappointed by their own work and have nothing to say to anyone. Transformational creativity, the kind that changes what a field counts as good, depends on understanding why the existing rules mattered, and that is not yet within their reach.

None of this means the technology is unimportant; it means that the word creativity is being asked to carry an argument it cannot bear. The useful question is not whether a machine has imagination but what will happen to human judgement, taste and responsibility when competent novelty is cheap. My guess is that these qualities will become more valuable rather than less. Deciding what is worth making, and being answerable for the result, will remain human work.