![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Ten years ago today I walked into the lobby. I got my photo taken and picked up my badge and a laptop. Or, as we say at Google, today is my tenth Googleversary.
Less than a month later Google's security team discovered Operation Aurora, an advanced cyberattack campaign conducted by China aimed at obtaining intellectual property from U.S. companies and, in Google's case, email data from Chinese dissidents. Before announcing this attack, Google instituted some fairly disruptive security requirements—Googlers around in 2010 probably remember entering OTP codes seemingly every time they wanted to interact with an internal resource. One of the things I've really admired about Google is that the security team didn't stop with a secure-but-onerous solution. They recognize that security and usability don't have to be in tension, and over the last decades internal security has increased while simultaneously being easier to use.
I've often felt like I joined Google at an inflection point in its history, and the history of the Internet. The Operation Aurora attack was one symbol of it: user data was quickly becoming the most valuable piece of the Internet. The early 2010s saw the decline of many open protocols—from CreativeCommons to OpenID to RSS to preventing Facebook from scraping contact data—as companies focused on capturing users in walled gardens and extracting value from user data. Internally, Google was adapting the scale challenges of tens of thousands of employees (in the early 2000s they'd shifted from hundreds to thousands and could no longer all be in the same place at once) and not long after reorganized from parallel org charts for engineering, product, and user experience to an org chart where each product-area was its own tree.
Google was also assessing its product strategy: for ten years they'd launched a lot of experimental products and features. Some had become wildly successful, like GMail and Maps, while others hung out in something of an unfinished state, not getting the investment to become truly useful. The Internet was shifting no longer was it email plus a world wide web of random interesting stuff. The iPhone was fairly new, and on my second week at Google everyone in the company got an Android Nexus One as a holiday gift. Social connections were also clearly a new important Internet dynamic, and Google invested heavily in addressing that gap.
User data, mobile, social, and security. And cloud computing. That's a good pass at the Internet themes of the past decade. Throw in some machine learning, because there's finally enough compute power to scale those algorithms. And while I wasn't at Google for the first ten years, it felt like I joined at the time of a big shift to play in this new sandbox.
Today, I think it's pretty clear that both the Internet and Google are undergoing another inflection point. Internally, Google's feeling the scaling crunch of a shit from tens of thousands of employees to hundreds of thousands. At quarterly engineering meetings we used to see detailed presentations on Search and Ads, and even folks who didn't work on those products were pretty well versed in their domain models. Today a meeting with a deep dive on every billion-dollar business would take all day, and most engineers don't have the context to understand everything that's going on in another product area. Google's also struggling with symptoms that are visible on the Internet at large: hostility towards perceived out-groups, lack of trust in institutions and processes, and a high-stakes focus on controlling the narrative.
Today when I hear about something on the Internet I'm a lot likely to say "ugh!" rather than "awesome!" which is a big shame. In the aught naughts and in the nineties the ratio was high on awesome, and a big slice of the ugh was for stuff that now feels trivial, like under-construction gifs and garish MySpace layouts.
Less than a month later Google's security team discovered Operation Aurora, an advanced cyberattack campaign conducted by China aimed at obtaining intellectual property from U.S. companies and, in Google's case, email data from Chinese dissidents. Before announcing this attack, Google instituted some fairly disruptive security requirements—Googlers around in 2010 probably remember entering OTP codes seemingly every time they wanted to interact with an internal resource. One of the things I've really admired about Google is that the security team didn't stop with a secure-but-onerous solution. They recognize that security and usability don't have to be in tension, and over the last decades internal security has increased while simultaneously being easier to use.
I've often felt like I joined Google at an inflection point in its history, and the history of the Internet. The Operation Aurora attack was one symbol of it: user data was quickly becoming the most valuable piece of the Internet. The early 2010s saw the decline of many open protocols—from CreativeCommons to OpenID to RSS to preventing Facebook from scraping contact data—as companies focused on capturing users in walled gardens and extracting value from user data. Internally, Google was adapting the scale challenges of tens of thousands of employees (in the early 2000s they'd shifted from hundreds to thousands and could no longer all be in the same place at once) and not long after reorganized from parallel org charts for engineering, product, and user experience to an org chart where each product-area was its own tree.
Google was also assessing its product strategy: for ten years they'd launched a lot of experimental products and features. Some had become wildly successful, like GMail and Maps, while others hung out in something of an unfinished state, not getting the investment to become truly useful. The Internet was shifting no longer was it email plus a world wide web of random interesting stuff. The iPhone was fairly new, and on my second week at Google everyone in the company got an Android Nexus One as a holiday gift. Social connections were also clearly a new important Internet dynamic, and Google invested heavily in addressing that gap.
User data, mobile, social, and security. And cloud computing. That's a good pass at the Internet themes of the past decade. Throw in some machine learning, because there's finally enough compute power to scale those algorithms. And while I wasn't at Google for the first ten years, it felt like I joined at the time of a big shift to play in this new sandbox.
Today, I think it's pretty clear that both the Internet and Google are undergoing another inflection point. Internally, Google's feeling the scaling crunch of a shit from tens of thousands of employees to hundreds of thousands. At quarterly engineering meetings we used to see detailed presentations on Search and Ads, and even folks who didn't work on those products were pretty well versed in their domain models. Today a meeting with a deep dive on every billion-dollar business would take all day, and most engineers don't have the context to understand everything that's going on in another product area. Google's also struggling with symptoms that are visible on the Internet at large: hostility towards perceived out-groups, lack of trust in institutions and processes, and a high-stakes focus on controlling the narrative.
Today when I hear about something on the Internet I'm a lot likely to say "ugh!" rather than "awesome!" which is a big shame. In the aught naughts and in the nineties the ratio was high on awesome, and a big slice of the ugh was for stuff that now feels trivial, like under-construction gifs and garish MySpace layouts.
