20 years of Google: Privacy, fake news and the future

The article by Rachel Lopez was published in Hindustan Times on August 26, 2018. Pranesh Prakash was quoted.

Happy Birthday, Google. The search engine is 20 this year, and what a ride it’s been! When Sergey Brin and Larry Page were developing software that searched better and loaded faster than Explorer, Navigator and AltaVista, the web itself consisted of just 1 lakh websites.

Google’s mission statement was succinct: To organise the world’s information and make it universally accessible. Their corporate code of conduct was even simpler: Don’t be evil.

Perhaps even Google didn’t realise where its mission would take it. The following decade brought Google News, Gmail, Maps and Chrome. By 2014, the internet had grown to 1 billion websites. The search engine, their core product, had become the default homepage of the Internet.

In May this year, Google quietly dropped the ‘Don’t be evil’ tag. The same month, its Android operating system crossed 2 billion monthly active devices. Seven products (including YouTube and Google Play) now reach a combined 1 billion users.

Google once directed you to information. Today, it’s often the source of information (in ads and top-of-the-page blocs), using data you and others have shared, often without you realising it. Public knowledge goes where Google takes it. And 20 years on, not everyone’s happy with the journey.

“The key concern is that Google has grown so big,” says Pranesh Prakash, policy director at Bangalore’s Centre for Internet & Society. “It’s like the classic line from [Spiderman’s] Uncle Ben: With great power comes great responsibility. In Google’s case, its great size is what brought great power to begin with.”

For billions of Google users, the biggest concerns are now of privacy and accountability, says Nikhil Pahwa, founder of Medianama, which analyses digital and telecom businesses. “There are few checks on Google’s ability to take, retain and process information from users,” he says.

Hits and misses

For Google, all is going according to plan. Its search engine is now smart enough to complete your sentences. It’s learning constantly from what you search for, watch, spend on, share and regret; it knows your commute and your vacation plans. And it’s profiting from this knowledge.

In the UK, Google is being sued for bypassing iPhone privacy settings to track and collect data from 4.4 million users in 2011 and 2012. Information on race, physical and mental health, political leanings, sexuality, shopping habits and locations was apparently used to build advertising categories. Google also creates products for the US government, and has user data from around the world. “Any entity that has this much insight into us, and is in a position to use it, whether for the government or commercial gain, is cause for worry,” says Prakash. Most users aren’t worried, and that’s worrying too. We don’t realise how much data is being tracked or collected. The more we share, the more useful Google gets, and the greater its potential for misuse, for mapping say, beef-eaters, online dissenters, LGBT supporters or single women who work late.

The Internet’s other giant, Facebook, recently suspended 400 apps over privacy concerns, admitting that 87 million users may have had data compromised in 2016. Meanwhile, even non-Google apps are capable of hijacking data using software developed by Google. Weather apps look at your photo gallery, ride-sharing software keep tracking you after the ride, games are checking out your texts as you play. Gmail knows your flight timings, how many steps you’ve walked, and your last bank transaction.

Search for tomorrow

Perhaps the biggest concerns are with Google’s artificial intelligence technology, the brand’s great leap forward fuelled by its massive data reserves. The tech is already being criticised for being fed biased data, creating global services that mirror the prejudices of an insular, mostly white, mostly male, tech industry.

Sara Wachter-Boettcher, author of Technically Wrong, which looks at how technology reflects sexism and the biases of the people that create it, says this creates problems. “Google develops tools that other tech companies rely on to build other products,” she says. So its biases spread to other products too. As machines learn, Google is starting to unlearn too.

“Machine unlearning is basically recognising when a machine has learned something inaccurate, or biased, and then erasing that learning,” says Wachter-Boettcher. In Africa, the company (along with Facebook) now funds a Masters course in machine intelligence to improve the industry’s diversity. Last year, Google took its first steps to curb fake news hits on its search engines with tools that allow users to report misleading or offensive content.

But perhaps it’s time to work towards a future in which Google will be monitored in real time, in different countries, rather than depending on the company to offer a fix after a misstep. Prakash believes that the way forward is reimagining an Internet where Google isn’t the first and last word on everything. “This doesn’t mean more companies like Google but searching that happens in a more decentralised way,” he says. “We need to save the web from large monopolies in the long run.”