IELTS Academic Reading Practice Test 137

In March 1989 a British software engineer named Tim Berners-Lee, then working at CERN, the European particle-physics laboratory near Geneva, gave his manager a document setting out a scheme for managing information. CERN employed thousands of visiting scientists who arrived with incompatible computers and their own preferred software, and valuable files were regularly lost when researchers went home. Berners-Lee proposed that documents should be linked to one another by references written into the text itself, so that a reader could move from one file to a related file stored on any machine on the laboratory's network. His manager, Mike Sendall, wrote on the cover what has become a celebrated comment: 'Vague but exciting'.

The idea of linking documents in this way was not new. In 1945 the American engineer Vannevar Bush had imagined a desk-sized machine, which he called the Memex, in which an individual could store books and records and create trails between them. In 1965 the sociologist Ted Nelson coined the word 'hypertext' for text that branches and allows the reader to choose a path, and in 1968 Douglas Engelbart demonstrated a computer system in which clicking on a link opened another document. Meanwhile the internet itself, a network of networks communicating through a common protocol, was growing. Berners-Lee's achievement was to combine these strands.

To make the scheme work he had to invent three things, each of which is still in use. The first was a simple language for marking up a document, now known as HTML, which tells a browser how text should be arranged and where the links are. The second was a set of rules by which a program asks a server for a file and the server replies, called HTTP.

The third was a system of addresses, the URL, that identifies every document and every other resource on the network. A URL combines the name of the server holding a file with the path to that file, an arrangement that allowed any computer to find any other without being told in advance where to look.

By the end of 1990 Berners-Lee had written the first browser, a program that could both display a page and edit it, running on a NeXT workstation, and had set up the first web server on the same machine. The first page it served described the project itself and explained how others could join in. A Belgian engineer at CERN, Robert Cailliau, became his closest collaborator and helped to persuade the laboratory to support the work. The name was debated: Berners-Lee considered alternatives such as 'Information Mesh' and 'Mine of Information' before settling on the World Wide Web.

The software was released to other users in 1991, and on 6 August of that year the first public website went live. Pages were plain text only. A second, simpler browser written by a student named Nicola Pellow could run on almost any computer, including the inexpensive terminals common in universities, and this did much to spread the idea. Progress was nevertheless slow at first: outside a small circle of physicists, few people had heard of the web, and those who had often assumed it was a minor feature of the internet rather than a new way of using it.

The decisive change came in 1993. On 30 April CERN issued a document placing the underlying code in the public domain, allowing anyone to use and modify it without payment, a decision Berners-Lee has repeatedly described as essential to the web's success. In the same year the National Center for Supercomputing Applications at the University of Illinois released Mosaic, a browser that displayed pictures alongside text on the page rather than in a separate window. Mosaic was easy to install and soon attracted hundreds of thousands of users, and the growth of the web became explosive.

Numbers illustrate the speed of the transformation. The web consisted of a single site in 1991; within five years the number of sites had reached several hundred thousand, and addresses ending in .com had become familiar in advertising and on television. In 1994 Berners-Lee left CERN for the Massachusetts Institute of Technology, where he founded the World Wide Web Consortium, an organisation that develops common standards so that pages written for one browser will also work in another. The first international conference devoted to the web had been held in Geneva that same year, with a few hundred participants.

Berners-Lee was knighted in 2004 and received the Turing Award, the most prestigious prize in computing, in 2016. He has since devoted much of his time to campaigning for an open web, arguing that a handful of companies now control too much of the world's online activity and that questions of privacy and access deserve as much attention as the original technical problems. In 2009 he founded the World Wide Web Foundation to promote that cause.

Yet the design decisions of 1989–91 have proved remarkably durable. HTML has been rewritten and extended many times, browsers have become vastly more powerful, and the machines that store pages have multiplied beyond anything CERN could have imagined. The essential principles, however, remain: a document is identified by a unique address, it is transferred by a simple set of rules, and it may link to any other document anywhere. The web is, in that sense, still the system described in a memo written in March 1989.

A single field is nowhere quite uniform. One corner may hold water after rain, another may be thin and stony, and the nutrients left by the previous crop will differ from metre to metre. Traditional farming ignores most of this: seed, fertiliser and spray are applied at one rate across the whole field, on the reasonable but crude assumption that an average is good enough. Precision agriculture measures the differences and treats each small area according to what it needs, an approach that became practical only once satellite positioning allowed a machine's exact position to be recorded.

The idea took shape in the late 1980s, when receivers able to fix a position to within a few metres became affordable. Accuracy improved again in 2000, when the United States stopped deliberately degrading the civilian signal from its Global Positioning System. From then on a farmer could attach a receiver to a tractor and know exactly which strip of which field was being treated. In Europe and North America the technology spread first on large arable farms, where the cost of the equipment could be justified by the area over which it was used.

Yield monitors, fitted to combine harvesters from the 1990s onwards, produce the maps on which much of the rest depends. A mass-flow sensor measures the quantity of grain passing through the machine, while a moisture sensor records how damp it is, since wet grain weighs more than dry. A GPS receiver notes the position of every reading, and software assembles the results into a map showing how productivity varies from one part of a field to another. Patterns quickly become visible that the farmer had never suspected.

Much of the information comes from the way plants reflect light. Chlorophyll absorbs strongly in the red and blue parts of the spectrum and reflects green, which is why leaves look green; in the near-infrared, healthy leaves reflect far more light than they absorb, while bare soil reflects comparatively little. A camera measures reflected light in the red and near-infrared bands, and software combines the two readings into a vegetation index, commonly NDVI, whose value rises with the amount of living leaf. The two Sentinel-2 satellites, launched in 2015 and 2017, image the land every five days at a resolution of ten metres, fine enough to show variation within a large field. Cloud is a persistent difficulty: in northern Europe a field may be visible from space only a few times between spring and harvest, which is why aircraft and drones carrying similar cameras are also used.

Where imagery shows a zone to be short of nutrient, the farmer can adjust the rate at which fertiliser is spread. Modern spreaders follow a prescription map, opening and closing sections or changing the dose as the machine crosses the field. Nitrogen matters most, because it is expensive and because the surplus does not stay where it is put. Some is washed through the soil into streams and groundwater, a process known as leaching, and some is converted by soil bacteria into nitrous oxide, a greenhouse gas roughly 300 times more powerful than carbon dioxide. Applying less where it is not needed saves money and reduces pollution.

Chemical sprays can be cut in a different way. A conventional boom sprays the whole field, including the square metres between weeds. Machines fitted with optical sensors, commercially available since the 1990s, shine light onto the ground and open a nozzle only when they detect the green of a plant against the brown of soil, and the saving can be substantial where weeds are sparse. More recent machines use cameras and machine-learning software to recognise particular species, so that a weed can be killed with a specific chemical while the crop is spared. The technique works best when crop and weed look different; separating two broad-leaved plants in a dense canopy is much harder.

Water is managed in much the same spirit. Instead of irrigating on a fixed timetable, growers bury sensors that measure how much water the soil holds and irrigate only when the reading approaches the level at which the crop will begin to suffer. Most such devices work by measuring the capacitance, or electrical storage capacity, of the surrounding soil, which changes with its moisture content; a related approach measures the tension with which soil holds on to water, using an instrument called a tensiometer. Because most of the water a crop takes up is lost through its leaves in transpiration rather than stored, precise scheduling can reduce consumption considerably, which matters most where water is scarce.

Sensors rarely measure what farmers actually want to know. A device recording the electrical conductivity of soil is responding to its texture, its salinity, its water content and its temperature at once, so the reading is a proxy that must be interpreted with care and checked against samples dug from the ground. Instruments drift and need calibration, and the cost of the equipment and of the software that makes sense of the data is considerable. Small farms, which dominate agriculture in most of the world, struggle to justify it, and mobile networks in rural areas are often too slow to carry the volumes of data involved.

Work now focuses on making the tools cheaper and simpler. Engineers are developing printed sensors that could be scattered across a field and read by a passing vehicle, and small robots that will measure and treat individual plants rather than whole zones. Satellite constellations are increasing the frequency and sharpness of images, while more powerful processors allow much of the analysis to happen on the machine itself. The hardest problem is not hardware but advice: a map is of little use to a farmer with no agronomist to explain it, and extension services in many countries are short of staff.

In parts of northern Europe the disappearance of cash is already far advanced. Buses in several Scandinavian cities no longer accept notes and coins, some cafés refuse them, and a growing share of the population says it rarely handles physical money at all. Britain and other countries are following a similar path more slowly, and the word used for the destination is the cashless society. That such a society is technically possible is no longer in doubt. Whether it would be a good one is a different question, and it is only recently that governments have begun to address it.

The case for digital payment is easy to state. A card or a phone can be read in a second, whereas handling notes takes time, requires a float of change and ends with somebody carrying the takings to a bank. Electronic records make accounting and tax collection simpler, and a shop that accepts no cash holds nothing worth stealing. Consumers value the convenience, and central banks have in general welcomed the shift, since coins and notes are expensive to print, to move around the country and to replace.

It is worth asking who pays for this convenience. Every card payment carries an interchange charge, a small percentage that the merchant's bank passes to the bank that issued the customer's card, and together with the fees charged by the payment networks these amount to a real cost of doing business. Cash is not free either, since it must be counted, insured and transported. The difference is that the cost of the card system is invisible to the customer, who is rewarded with air miles and cashback, while the cost of handling notes falls directly on the shop.

Then there is exclusion. Everyone who can obtain a bank account and use a smartphone can take part in a cashless economy, but not everyone can. Older people who have never used a card, people with learning difficulties, those who have lost their homes and victims of domestic abuse whose spending is watched by a partner all depend on notes and coins. Rural areas with unreliable mobile coverage pose a further problem, since a payment terminal is useless without a signal. Surveys consistently show that people on the lowest incomes use cash most, and whether the state has an obligation to keep a payment method for them is a political question, but it is not an empty one.

A cashless economy also concentrates risk. Notes continue to work when the electricity fails; a payment system does not. A fault at a single data centre, a faulty software update or a deliberate attack on a payment network can stop trade within minutes, as several countries have discovered during outages in recent years. Cash is, in effect, a backup that requires no power, no network and no permission. Its very inefficiency is part of what makes it resilient.

Privacy is the third consideration. Every electronic payment produces a record: what was bought, where, when and by whom. That record is useful to tax authorities and valuable to companies that wish to target advertising, and it is available to anyone who can obtain it, whether through a court order, a data breach or the routine purchase of commercial databases. The law protects it imperfectly, and a person who pays by phone has, in practice, consented to a remarkable amount of observation. Notes leave no such trace, which is why they are favoured by criminals and by anyone with reason to value discretion.

Claims about crime deserve scepticism. The number of bank robberies in Sweden has fallen sharply, and it is often said that this proves the benefits of going cashless; but the fall coincided with the closure of rural branches and with better security, so the cause is not obvious. Fraud and money laundering, meanwhile, have not declined: they have moved online, where transactions are faster and borders mean nothing. Removing cash removes one kind of crime and enlarges another.

Two examples from outside Europe show how much depends on design. In Kenya, the mobile money service M-Pesa, launched in 2007, allowed people without a bank account to send and store value on a basic phone, and it is widely credited with extending financial inclusion. India's decision in November 2016 to withdraw the 500 and 1,000 rupee notes, by contrast, removed the great majority of the currency in circulation almost overnight, and the inconvenience fell most heavily on people who had no alternative. Digital payments grew, but so did hardship.

What follows from all this? Not, in my view, that the cashless society should be resisted. Digital payments are convenient, and for most people most of the time they are preferable to a pocket full of coins. What follows is that cash should be understood as a piece of public infrastructure rather than a private product, and protected as such: retailers selling essential goods should be obliged to accept it, withdrawal machines should be available within a reasonable distance, and any central bank digital currency should be designed to coexist with notes rather than replace them. The aim should be a system in which cash is rarely needed but always available — a backstop, not a competitor.