Don't be evil

Books A new book outlining Google's gargantuan ambitions unnerves Michael Brotzman, a former Googler.

Courtesy www.stephencollinsillustration.com

REF rv07no-brotzmangoogle 07/11/08
Powered by automated translation

A new book outlining Google's gargantuan ambitions unnerves Michael Brotzman, a former Googler.
Planet Google: One Company's Audacious Plan to Organise Everything We Know Randall Stross Atlantic Books Dh100
I've known Google for about seven years now, and I've noticed that our interactions always follow a similar pattern. Maybe you know it: Google offers me something convenient - something a little too convenient, it seems - and I accept it with a tight-lipped nod and sideways glance, unsure if I should be concerned about shaking hands with a company that is monstrous in size and perhaps (who knows?) motivation. These interactions always conclude with Google gaining a little more control over my life, and by now Google has a pretty firm grip on my daily interactions: browsing the web, checking e-mail, making weekend plans, chatting with friends - a lot of what I do nowadays is done through Google.

I never put much thought into formalising this awkward-handshake feeling until this past June, when I stumbled on a name for it while (ironically enough) shaking a hand. I spent the summer working for Google as a software engineering intern, and on my first day, I met an old friend-turned-co-worker for lunch. He greeted me with a handshake and the disarmingly casual "Welcome to Tentacle Command." I immediately went through the familiar motions (tight-lipped nod, sideways glance) as though I had actually performed them countless times. I churned the phrase around in my mind as we ate and then for the next few weeks, and I came to realise that it perfectly summarises what Google is about, what it's after and what it's like to interact with.

First there's "Welcome to", which implies a firm handshake with the entire Google culture, something I experienced daily during my interactions with "Googlers": votaries of the creepily insular culture (emphasis on "cult") of Google employees who all hang out with each other and all employ the same even, slow-up-and-down-handshake-maintain-eye-contact tone of voice when talking about their work life. I had spent the preceding four years watching friend after friend attain Googlership. I'm proud to say that I never did.

Then there's "Tentacle Command" - that cropped up everywhere. I remember listening to a senior engineer explain that Google would rather spend hundreds of thousands of dollars to purchase (or develop an in-house facsimile of) an attractive technology than pay $10 to license its use. I remember, during the weekly all-hands meetings I made a point of attending, watching Google's hardware gurus click through slide after slide of pictures from the new server farms Google was rolling out around the world, each a man-made canyon reminiscent of midtown Manhattan. A hungry organism, sliding out a tendril, constricting a target, snatching it back for digestion. This was the Google way: relentless expansionism.

I made giggly asides like these to myself throughout the summer, and was fully prepared to consign them to the back of my mind until I read Randall Stross's new book, Planet Google: One Company's Audacious Plan to Organise Everything We Know. My impressions had been, well, impressions: one-sided, subjective, anecdotal. Planet Google confirms my every conjecture (and more) with hard evidence. Stross is a journalist (he writes about technology for The New York Times), not a polemicist. He lets the facts speak for themselves, and his explicit conclusions are fairly limited as a result. Nevertheless, his book should serve as the starting point for any intelligent conversation about what Google is up to and why it matters.

Most of what Stross gives us is a series of (surprisingly gripping) biographies of Google and its products. Google was incorporated in 1998 by Sergey Brin and Larry Page (then doctoral candidates at Stanford University) based on research they were conducting into methods for finding and retrieving information on the web. Fuelled by good press, word of mouth advertising and the irresistible, down-to-earth appeal of its spartan aesthetic, Google grew steadily and easily survived the burst of the dot-com bubble around the turn of the century. In 2000, it launched AdWords, an advertising programme that took off quickly and propelled the company to a massive IPO in 2004. Since then, Google has continuously expanded and now employs more than 19,000 people in over 60 locations around the globe. It is the king of online search and ads: Google currently serves about 70 per cent of all web searches. In 2007 it earned $16.6 billion, 98.9 per cent of which was generated from advertisements.

As Google grew, it expanded its product line, and Stross devotes a chapter to each of its most prominent achievements outside of Search. There's Books, the ongoing attempt to digitise every book ever published, which has already claimed some two million titles and might add a million more this year. Then there's YouTube, the most popular video sharing site on the internet (streaming about three billion videos a month), which Google acquired in 2006 for $1.65 billion and which has yet to turn a profit. Google Maps and Google Earth let users poke around annotated satellite pictures of most of the planet (to the chagrin of sunbathers everywhere). Gmail, Google's revolutionary web-based e-mail client, set a new standard for its field when it debuted in 2004 by offering vastly more storage space than any other competitor. Its capacity and user base (already tens of millions strong) continue to expand today.

Stross also discusses the underlying philosophy that underlies every decision Google makes. It boils down to "keep your eye on the prize of scale at all times." All business moves Google makes are designed to accommodate a billion or more users. For example, Google uses lots of cheap, unreliable hardware systems because computation on Google's scale requires tremendous computing power and reliable hardware is too expensive to buy at Google's scale. Instead, Google uses software that accounts for hardware failures and constant upscaling. Its hardware development compliments (and facilitates) its software development. Every project Google undertakes is planned in terms of the maximum case: Books wants every book, Search every website, Images every picture, Maps every road and so on.

Stross concludes Planet Google by quoting Eric Schmidt, Google's CEO, who says that there is a countable amount of information in the world, which he and his colleagues have estimated it should take no more than 300 years to completely organise. Stross then suggests that 300 years may be an overestimate, implying that he believes Google can and will achieve its goal. It's a chilling claim, particularly since, coming on the heels of 200 pages of journalistic objectivity, it doesn't seem unreasonable. Closing and flipping over the book, you notice the subtitle again: One Company's Audacious Plan to Organise Everything We Know. It strikes you that "audacious" is the only tendentious aspect of the whole book. Is Stross right? Is Google really "audacious"? Everything you have just read is suddenly cast in a new light, and anxiety begins to take shape. It strikes you: wait, what exactly is Google's "plan", again? Organise everything we know? What does that mean? And should I be worried about it?

Google's plan is, for Stross, evident: the company is campaigning for a position no less impressive than Gatekeeper to (Literally) All The World's Information. It's hard to disagree. Google is continuously expanding: expanding the scope of its algorithms to encompass ever more plentiful and diverse media, expanding its user base, expanding its cultural impact, even expanding its physical presence by continuously unveiling larger and larger server farms around the world. Its mission statement is "to organise the world's information and make it universally accessible and useful." It's not a particularly modest goal: "the world's information" can be interpreted as liberally as you like, from "all of human congress" on down, and Google assumes a very liberal interpretation indeed. Eric Schmidt once said that "If you think about it, all the world's information includes personal information." Craig Silverstein, Google's director of technology, once reiterated Schmidt's 300-year deadline, adding that "the goal is to make computers smart enough" to understand "emotions and other nonfactual information."

But Google's mission statement does more than demarcate its ambition. It identifies Google as fundamentally different than other free-market companies, and subject to different concerns. By aspiring to make "the world's information universally accessible and useful", Google identifies itself as a broker in what might be loosely called the realm of information: it sorts and distributes information and facilitates discourse. Google has already achieved significant control over this realm ("to google" is now a dictionary-sanctioned word meaning "to search for information on the internet using the Google search engine"). Soon Google may achieve a "monopoly", though it's hard to say how that might be quantified - say, a large number of people relying on Google for a large number of their information transactions (finding information, checking e-mail, organising budgets, doing taxes, talking on the phone, telling stories). And this is indisputably Google's goal. (Favourite quote: "Perhaps in the future, we can attach a little version of Google that you just plug into your brain" - Sergey Brin.)

The realm of information is not an ordinary service market like, say, the shaving cream market. If the shaving cream market goes wrong - if shaving cream becomes expensive or scarce - we get ripped off, or we just shave less. But that's not true of the realm of information, which plays a more central role in political and cultural life: if our means of finding information deteriorate, arguments become less well-informed or more misguided (also more difficult to discredit, as our means of fact-checking suffer); if our venues of discourse lessen in quality or disappear, our conversations become less frequent and more muddied; if the language we employ and encounter becomes restricted, the expressiveness of our conversations diminishes too. Flaws like these are not just scary in the abstract sense, nor are they new - their destructive potential has been demonstrated in the recent past, as when the American news media failed to facilitate a clear, well-informed discussion about the build-up to the Iraq war.

So, should you be worried about Google's plan? Maybe you should. Google's technologies occupy increasingly significant positions in the realm of information. What if, for example, Search's output became somehow ineffectual, biased, manipulated or restricted? What if Google Books, which is well on its way to possessing most books in existence, someday (perhaps in conjunction with a low-powered hand-held reading device - the gBook?) makes significant encroachments on book distribution and the act of reading - and then becomes restricted? What will be lost to those who rely on Google for the kind of knowledge that comes only from books?

Google commands a position that is uniquely conducive to flaws like these. It is successful because it is big, not the other way around. The effectiveness of the services that it provides requires a huge amount of server power, a tremendous amount of data and the collaboration of an army of the smartest people in the world. For a new company to compete with Google, it will need (barring some unforeseen technological breakthrough) to start small and grow quickly like Google did, offer something that Google doesn't (no small feat) and fend off lucrative buyout offers along the way. And because the quality of an online information provider relies on scale, few people will have reason to switch from Google to a small competitor. The incredibly high cost of entry into Google's market affords it a lot of room to be inefficient someday. This potential is more than enough cause for concern; wild-eyed speculation about Googley dystopias is unnecessary and besides the point.

If Google gains an information monopoly and becomes seriously flawed, there will always be holdout forums: books, letters, everyday communication, the rest of the internet not yet colonised by Google. But millions of people will be increasingly enmeshed in a system that does not serve them or their societies particularly well, and the holdouts will have little chance of effecting change against the vast, entrenched and convenient status quo. For many, Google will be both the problem and the main means of discussing and addressing the problem.

And this is not a bad 50s B movie. If Google threatens the realm of information, there will be no dramatic solution (eg Sergey Brin, clutching a tear-stained photo of Larry Page as he weaves his spaceship through the gTentacles surrounding Tentacle Command, ignores the pleas of some bespectacled Googler that sputter through the on-board intercom system: "No, Sergey - don't do it!" But it's too late. Sergey drains his afterburners, slips through the gMaw and smashes the glass on the shiny red switch - H-bomb activated. Mission accomplished.) Instead, our public discourse will simply wane and falter a little bit more as those under Google's sway become increasingly circumscribed.

Whether or not this actually happens, it is the logical entailment of Google's plan, which - well, come to think of it, is "audacious" after all: "ambitious", check; "bold", "innovative", check, check; and "insolent," check. In so organising information, Google disrespects its very object.
Michael Brotzman is a graduate student in computer science at Columbia University.