Thoughts on Google

As I watched the Google I/O keynote yesterday, I realized once again that I can’t make up my mind whether I should love or hate Google. In fact, there is probably no other tech company I have more conflicting feelings about.

Google is Amazing

In web search, Google remains the best.1 Google Maps is one of my favorite products of all time, and continues to delight me almost every time I use it. The same is true for Earth and Street View. YouTube is a bottomless treasure trove of entertainment and education. Even after ten years, Gmail is still the best webmail system on the market in a lot of ways (though I no longer use it due to privacy concerns). I also truly admire Google Books and the company’s research in self-driving cars, robots, and virtual reality.

The new Google Photos looks very impressive. Given the company’s expertise, I have no doubt the algorithms to automatically tag and organize your photos will work great. I am convinced Google is the best company in the world at this task, probably way ahead of anything Apple could do.

All these products are as useful as they are technically impressive, and I have a lot of respect for the people who build them.

Google’s mission statement is to organize the world’s information and make it universally accessible and useful. It’s an admirable vision, and I do believe Larry Page and Sergey Brin mean it sincerely.

Google is Evil

Unfortunately, in Google’s mind, “the world’s information” also includes everything there is to know about you and me: our private messages, our browsing habits, our location, our preferences, our friends, and possibly much more. On the web, the widespread use of Google Analytics, Google Fonts, Google-controlled ads, and Google-hosted Javascript libraries mean that Google can probably track you across a majority of the pages you visit.2 (And by using these services on our sites, we website owners are complicit in this scheme.)

Many iOS apps also use Google Analytics or Google’s ad services. And if you have an Android phone, Google obviously knows even more about you. I’ve always been an iOS user, and the main reason for me not to switch to Android used to be usability and the availability of apps. Nowadays it is that I don’t want Google to have control over my smartphone.

Google says they currently don’t use the stuff they learn about you from your photos in other products, but I’m sure that will change. I imagine that soon you won’t have to tell Google anymore that you’re friends with someone because it can easily tell that from your and your friend’s photostream.

I don’t know about you, but I find that incredibly fascinating and incredibly creepy at the same time.

Two Kinds of the Web

I realize that I have to give up some privacy to use any cloud service (from Google or any other company). I would like to see services that allowed me to permit them what they can learn from the things I share with them, and I increasingly feel that I don’t get that kind of control.3 I trust Apple much more than Google in this regard because Apple clearly has a business model that is not based on selling my information.

Bret Victor recently argued that the web really consists of two kinds of information: firstly, the “common record”, a formal body of published art, knowledge, and commentary. Secondly, a “conversational record” of ephemeral data that should not be permanent:

Photos from your friend’s party are not part of the common record.

Nor are most casual conversations. Nor are search histories, commercial transactions, “friend networks”, or most things that might be labeled “personal data”. These are not deliberate publications like a bound book; they are not intended to be lasting contributions to the public discourse.

The fact that they can persist — that the medium is constructed such that this data accidentally travels further and lasts longer than anyone intended, and that this data is easily retained and exploited by intermediaries — is an enormous and terrifying problem of its own.

Google has done a tremendous job in collecting and organizing the common record. Increasingly, it also has become interested in the conversational record.

I’d love a Google that only cared for one half of the web.

  1. Especially if you do searches in languages other than English, I have the impression that Google’s search quality is even further ahead. I can’t back this up with numbers, though. ↩︎

  2. Facebook does this too, of course. Unlike Google, it’s much easier for me to hate Facebook because I don’t seriously use any of their products. Though even that has become harder as I’ve gotten to know a lot of awesome people who work for Facebook and do fantastic open-source work in the iOS community. I hope it’s possible to respect the people who work for Facebook but not the company as a whole. ↩︎

  3. I think Microsoft’s Cortana is a good example of a service that gets this right. As far as I know, Cortana keeps a “notebook” of everything it knows about the user, who can then check and optionally delete information they don’t want to share. ↩︎