For years, Google tried to fight against its name becoming a generic verb. It failed, just like Hoover and Xerox before it. Searching the internet is so inextricably linked to Google, that Microsoft could spend all the money it’s got in the bank and the world would never start talking about “binging” things. But as Google tries ever harder to encourage us to ask more questions using our voices and share more detail so it can predict what we want before we even ask, the serach engine wants to move away from the simplicity of search –– turning it from an active verb into a passive one.
In a blog post published earlier this week, Ben Gomes, Google’s head of search, compares the web which Google indexed in 1998 to the equivalent of a small library, 25 million pages, and notes that the billions of pages of information it now understands could not be contained within all the libraries on earth. Besides making an unintentional argument for stopping all those library closures, his point is that Google has to keep expanding well beyond a search box and a pair of buttons. As Google tries to reach more and more people across the world — including areas with high levels of illiteracy or which lean more towards spoken than written communication — it makes sense that it’s focusing more on images and voice as ways to give people information, but its plans around AI and promises of more “serendipity” in search should make us pause.
Take, for example, a new feature that the company unveiled at its 20th anniversary celebration event on September 24 — Discover, a feed of videos and articles that Google predicts an individual user will find interesting, displayed on the Google home page, below the search bar in all mobile browsers. Initially that sounds useful and utterly harmless, but think about the ongoing problems of both misinformation and partial news. This is another step by Google to tighten the filter bubble around its users, presenting you not with a breadth of sources delivered from a search you chose to make, but a set of perspectives based on previous behaviour. It’s the search equivalent of the YouTube sidebar, which endlessly serves up similar videos to the ones you’ve already watched.
Underpinning Google’s increasing and seemingly magical ability to predict what you want to search for, when you want to search for it, is an ever-growing pool of data. In the beginning, Google simply had the search queries that users entered to work from, but with its expansion of services, more people are logged in, linking their searches to them personally and, through location data, where they go. And rather than treating each search as a discrete action, Google is increasingly treating search as “a journey”, tying your queries together. That’s presented as another handy feature for you, the user, but the people most interested in the journeys you take online are marketers — Google’s actual customers.
Another worrying aspect of Google’s developing approach to search is its intent to show more and more information directly in search results. That’s arguably great for users — they get what they’re looking for without having to click through to a different website — but terrible for anyone who did the original work to put information online. Google gets the benefit and the creator gets none of the traffic and no chance of building a relationship with the person who searched for it.
Gomes’ blog post includes this startling fact: “We see billions of queries every day, and 15 per cent of queries are ones we’ve never seen before.” But how will that number change as Google increasingly becomes less of a robot librarian you approach with a query and more of an information sommelier, pouring out the results it knows will be to your taste? Much has been written about search engines and apps becoming a form of extended cognition, requiring us to rely less on our memory, but how comfortable should we be with them taking on another step in the thought process — what actually interests us to start with?
The ever more serendipitous Google has to surveil us more to know what we want before we do. It also has the potential to create a feedback loop of ever decreasing value, as we become accustomed to the ease of consuming the information that it offers up to us, rather than choosing to search beyond our assumptions. More than ever, we should be questioning how much we are willing to give Google in return for convenience. Our human experiences should drive the algorithms, rather than be driven by them. Of course, Google can aid you in discovery, but it should remain a guide to navigating our questions and not turn into a guru.