In A Human’s Guide to Machine Intelligence, Kartik Hosdanagar explains how they “are shaping our lives and how we can stay in control” of our relationships with them.
He observes, “When you think of the word ‘algorithm,’ you might picture a computer crunching numbers according to a formula. But stated quite simply, an algorithm is merely a series of steps one follows to get something done. For example, I follow a series of steps when I make an omelet. You might call it an omelet recipe, but the former engineer in me views it as an omelet algorithm.”
The passage that follows immediately caught my eye as I began to read Chapter 5:
“At the heart of Google’s $700 billion [actually now $1.207 trillion] empier is its search engine, and at the heart of its search engine is its ranking algorithm. Back in 1999 Google cofounders Larry Page and Sergey Brin published a detailed description of their PageRank algorithm online. While it has some mathe4matical subtleties, PageRank is ultimately a fairly predictable algorithm: the more times your website is linked to from other websites, the higher it will be ranked in search results. However, because Google’s algorithm was both public and predictable, it created a perverse incentive for many website owners to generate ‘shadow’ websites whose sole purpose was to link back to their primary domain.
“As such, much of the engineering effort at Google in recent years has been devoted not only to improving the results of web-ranking algorithms but also to fighting spam and the work of search engine manipulators. After its initial stint of intellectual openness, Google has grown famously secretive about which factors affect a website’s position in its search results. More recently, its ranking algorithm has leaned heavily on machine learning principles, based on data about which results got the most engagement from previous searches.”
In the next chapter, Hosanagar discusses an idea that has gripped him for the past several years: “the notion that modern algorithms might be best understood through a lens from human psychology — in part because we humans are their creators.”
Humans may have created algorithms but will they be able to control them in months and years ahead? Opinions are divided.
* * *
It’s worth noting that early in 1999, when Page and Brin were Ph.D. candidates at Stanford University, they tried without success to sell their research project. It later became Google and then Alphabet Inc. Their asking price? About $750,000. Today it is worth about $1.2 trillion.
* * *
Kartik Hosanagar is the John C. Hower Professor of Technology and Digital Business and a Professor of Marketing at The Wharton School of the University of Pennsylvania. Kartik’s research work focuses on the digital economy, in particular the impact of analytics and algorithms on consumers and society, Internet media, Internet marketing and e-commerce. A Human’s Guide to Algorithms was published by Viking (March 2019).