When you last used your supermarket loyalty card, or pulled up just as the traffic lights turned red, you probably weren’t thinking much about math. Why would you? Most of us see mathematics as baffling school algebra that has no application in real life. But you probably meet hundreds of bits of complex math every day without realizing it: the discounts on your loyalty card are the result of deep data analysis, and those traffic signals run on an algorithm that determines when the lights will switch to green.
In fact, large bits of modern lives are secretly underpinned by complex algorithms — the mathematical equivalent of cookery recipes, which take a piece of information and turn it into an action or decision. Algorithms are like computer programs or flowcharts — a sequence of steps that examines what is happening and comes to a conclusion. Take traffic lights: The computer controlling them asks a series of related questions. What time of day is it? When did the lights last go red? Has a pedestrian pushed the button at the crossing? The algorithm guides the computer, step by step, to change the lights.
Algorithms are pervasive, even controlling, in our lives. Book a low-cost flight and an algorithm will determine how much the tickets will cost, depending on supply and demand; arrive at the airport and air-traffic algorithms will determine which place in the queue your plane gets. Watch the weather forecast on TV and the predictions will have been fine-tuned by an algorithm; listen to the radio and the playlist may have been generated by one. Supermarkets, in particular, are a hotbed of algorithms. Almost every aspect of their operation — from deciding the order that products are stacked on the shelves to picking which special offers to run — is determined by a computer.
And since computers are increasingly dominant in our lives, algorithms are increasingly important — and nowhere is this more apparent than on the internet. In the online world, mathematical analysis isn’t just important: the algorithm is king. Everywhere you turn online, companies are using algorithms in their quest for success. From Google’s search results and Apple’s music recommendations to Amazon telling you that “customers who bought this item also bought...” algorithms are at work.
“There is no way, with the size of the internet, that one can do exhaustive searches,” says Marcus du Sautoy, professor for the public understanding of science at Oxford University, and one of Britain’s leading mathematicians. “So you have to rely on mathematics to give you clever and fast ways to get information.”
The British online supermarket Ocado, for instance, analyzes every detail of every activity — from the choices shoppers make to warehouse movements — to make sure there is enough stock in the right places, and to help the company plan future strategies.
“We’re using complicated forecasting algorithms to predict demand,” says Jon Rudoe, head of retail. “The customer sees nothing. It all looks very simple, but it’s actually governed by complex mathematics in the background” — mathematics we can put to use because we now have technology that didn’t exist a decade ago.
Since these recipes have helped internet companies cook up billions in profits, the precise details of the mathematical mechanisms are jealously guarded as among the companies’ most valuable assets. Forget the recipe for Coke or the Colonel’s blend of herbs and spices, these are the trade secrets of the 21st century. And wherever there are secrets, there are people desperate to unlock them. Around the world, countless hours and millions — perhaps billions — of dollars are spent trying to unravel the inner workings of the web’s most powerful algorithms.
Among those searching for answers is Russell Davies, an advertising consultant from London. His fascination started when he decided to try to understand how to lift his book — a guide to Britain’s greasy-spoon cafes called Eggs, Bacon, Chips and Beans — up Amazon’s rankings. On the surface, the Amazon charts seem straightforward. They are largely based on sales, but there is also analysis of buying activity so that a sudden rush of orders can push a book, CD or DVD up the chart. The rewards for chart-topping are enormous: it can drive thousands more sales. But there are other benefits too, including (crucially) a shot at appearing in the algorithmically generated “customers who bought this also bought...” section, which is hugely popular. Davies thought he could try a variety of tricks to shift his book up the bestseller list and — if he succeeded — work out the secret ingredients behind Amazon’s recipe.
“I started to think about what makes the number change and realized that it’s a big secret, which made it more interesting,” he says.
Over the course of several months it became a minor obsession; after all, for writers and publishers, even a small insight into Amazon’s algorithm could be like discovering the Rosetta Stone.
“The Amazon ranking is the only feedback the average author gets on how their book is doing, and you’re desperate for feedback,” Davies says.
He wrote about the book in newspapers and online; he asked friends to buy copies and even bought a few himself, all the time closely monitoring whether his actions were reflected by his place in the rankings. But although he had a few moments when he seemed close to a breakthrough, he hasn’t cracked the code.
“I learned that you can’t really influence it,” he concludes. “Thinking about it, it’s not that surprising — you’re in such a massive pool of data that a few sales here or there are just invisible.”
Amazon is not alone in working hard to keep its methods confidential. EBay’s reputation system is largely based on user feedback, but the company constantly works on adaptations aimed at stopping scammers from getting status they don’t deserve. Apple, similarly, has mathematical secrets that are increasingly important to its business, among them a recent addition to the iPod, the new “Genius” function that creates playlists of similar songs.
On the surface, Genius looks like hi-tech wizardry. It takes a song you own and works out similar music that you might like to hear: Stick in a shiny happy song and you’ll get 25 tracks of sunshine back; give it a mournful dirge and you’ll end up with a mixtape to die to. Underneath all the whizziness, though, the Genius function is really about number crunching. Apple analyzes the song choices of millions of other iTunes users around the world and, based on this, is able to take a stab at which songs match your seed track. And it’s profitable, too: Genius can encourage users to hear and buy songs from iTunes that they knew nothing about.
Another aspect of Apple’s business it keeps quiet about is the way its iTunes charts are put together. While the charts appear largely based on sales and the number of times people listen, iTunes is also believed to use a so-called “decay algorithm” to give more weight to very recent activity. The company refuses to confirm how the system works, although little pieces of information have slipped out in the past. Its podcast chart, apparently, is “driven purely by an algorithm that looks at new subscriptions during the past week,” for example.
Nobbling the mathematics behind iTunes is almost impossible, but some people have learned how to subvert it the old-fashioned way. Little-known group Hit Masters rode high in the charts earlier this year with a karaoke cover of the recent Kid Rock hit All Summer Long. They managed a modern twist on the age-old tactic of riding somebody else’s success, when they noticed that the original version was not available to buy on the US iTunes store. Their speedily produced cover appeared high in the search for the Kid Rock song, making them more likely to get downloaded; more downloads meant a higher placing in the chart, which itself generated more sales. It’s a self-fulfilling cycle that eventually pushed Hit Masters to a chart high of 19 in the US Billboard Hot 100.
But the efforts of Amazon, eBay and Apple pale in comparison to the most famous algorithm on the Web: Google’s algorithm is the mathematical engine that drives the web’s most powerful company.
“It’s not that Google is smarter at the maths, it just had a better recipe,” says Danny Sullivan, editor of the SearchEngineLand.com Web site. “It had ingredients that other people didn’t.”
Given Google’s dominance of web search, getting a high ranking there is a guarantee of more clicks — and more clicks means more money. With so much at stake, decoding Google’s algorithm has become an industry in itself. Experts in search engine optimization know the basics of Google’s operations — they are well documented — but the California company now spends vast amounts of time and energy trying to keep its formula secret.
“Large sites such as Amazon and Google are tweaking the internals of their systems almost constantly,” says Jon Kleinberg, a professor of computer science at Cornell University in New York state. “The front end of Google looks the same to us, but behind the scenes they can be busily swapping out ideas.”
Kleinberg believes the internet is making so much data available about how we relate to each other that we could soon be able to create algorithms for the social interactions we always thought were too complicated for math, such as controversy, disagreement or fame.
“They’ve always been fleeting, ephemeral, invisible and essentially unknowable,” he says. “Now we can try to get in there and understand why, at a microscopic level.”
All of which will have potential applications, both corporate and individual. Your PC could use algorithms to recognize exactly what document you are looking for, or to predict which news stories you might be interested in when you log on. Your mobile phone could recognize you are in a bad mood and screen your calls automatically, allowing only people that it has determined are your closest friends to get through.
Learning to sift through the vast amount of information being sent across the internet every second to divine people’s feelings or intentions could, Kleinberg believes, be the next great technological leap. Mathematicians rule!
Two sets of economic data released last week by the Directorate-General of Budget, Accounting and Statistics (DGBAS) have drawn mixed reactions from the public: One on the nation’s economic performance in the first quarter of the year and the other on Taiwan’s household wealth distribution in 2021. GDP growth for the first quarter was faster than expected, at 6.51 percent year-on-year, an acceleration from the previous quarter’s 4.93 percent and higher than the agency’s February estimate of 5.92 percent. It was also the highest growth since the second quarter of 2021, when the economy expanded 8.07 percent, DGBAS data showed. The growth
In the intricate ballet of geopolitics, names signify more than mere identification: They embody history, culture and sovereignty. The recent decision by China to refer to Arunachal Pradesh as “Tsang Nan” or South Tibet, and to rename Tibet as “Xizang,” is a strategic move that extends beyond cartography into the realm of diplomatic signaling. This op-ed explores the implications of these actions and India’s potential response. Names are potent symbols in international relations, encapsulating the essence of a nation’s stance on territorial disputes. China’s choice to rename regions within Indian territory is not merely a linguistic exercise, but a symbolic assertion
More than seven months into the armed conflict in Gaza, the International Court of Justice ordered Israel to take “immediate and effective measures” to protect Palestinians in Gaza from the risk of genocide following a case brought by South Africa regarding Israel’s breaches of the 1948 Genocide Convention. The international community, including Amnesty International, called for an immediate ceasefire by all parties to prevent further loss of civilian lives and to ensure access to life-saving aid. Several protests have been organized around the world, including at the University of California Los Angeles (UCLA) and many other universities in the US.
In the 2022 book Danger Zone: The Coming Conflict with China, academics Hal Brands and Michael Beckley warned, against conventional wisdom, that it was not a rising China that the US and its allies had to fear, but a declining China. This is because “peaking powers” — nations at the peak of their relative power and staring over the precipice of decline — are particularly dangerous, as they might believe they only have a narrow window of opportunity to grab what they can before decline sets in, they said. The tailwinds that propelled China’s spectacular economic rise over the past