Rise of the google machine

Mon Jun 1, 2015

At Google I/O last week, developments in Android M (Milkshake? Macaroon?), Android Pay, Android Wear, Internet of Things, and VR all made an appearance.

However, it was Google’s advancements in machine learning that were the most impressive. The awkwardly-named “Now on Tap” in Android M exemplified these improvements nicely. Basically, Google will now offer in-context information within apps, emails, or web pages when pressing and holding the home button on your phone.

For example, if your receive a text from your wife asking you to pick up dinner, Google will offer to create a reminder. Any mention of movie titles will display links to the movie’s trailer, reviews, and other info. Restaurant names will trigger a formatted yelp review and OpenTable reservation options.

You could either marvel at the progress Google is making in artificial intelligence and machine learning, or you could be terrified by it. Either way, a healthy respect for this gargantuan digital overlord is probably wise.

But could respect ever blossom into trust? During Google I/O this year, it became clear that Google is courting us and wants to be the hub of our digital lives. It turns out that Google Plus was a misstep in that direction, so Google is trying again. From the detritus of its dismantled social network, comes the new Google Photos - a standalone version of the excellent photos management division of Google Plus.

The singular joy of Google Plus photo management was that Google offered unlimited photo and video uploads. However, photos were downsized considerably. This time around, Google is leveraging its incredible server farm assets to upload unlimited photos in their native resolution (up to a 16MP maximum).

This extremely generous cloud storage solution, coupled with seamless synchronization between a dedicated website and apps on both Android and iOS, makes Google Photos the most compelling photo management tool available today. Heck, I signed up on launch day and am still gobsmacked by how good it is.

But how far can you trust a company whose raison d’ĂȘtre is to sell targeted ads? I mean, photos are intensely personal - intimate snapshots of our family lives, even our secret lives. Are you comfortable with that relentless Google machine silently parsing every picture you hand it? Because that is exactly what is happening.

Google positioned its machine learning as a positive feature, and at face value it is amazing: Search for any term, such as “flowers” for example, and Google will show you all the flower pictures you’ve taken (even though you’ve never tagged them). Type in “flowers in San Diego” and Google will check geo-location data to only show you pictures of flowers you took on that trip to San Diego a few years ago. Crazy cool, right?

Google is trying to build trust here, so there is no advertising in Google Photos…yet. Targeted ads based on your photo collection does not seem far-fetched, however. In fact, it seems inevitable, especially once millions of users have invested the time and energy to curate their memories through Google’s “free” service.

Am I okay with this? Well, for the most part, I am. You’d be crazy to throw all your eggs into Google’s basket though. I am definitely keeping my photos stored locally as well. And it’s not as if I haven’t made forays into photo cloud storage. I tried Flickr’s redesigned photo service (offering a healthy 1 terabyte), but it’s just not as good. I didn’t even read Yahoo’s small print, which is probably horrifying.

The main reason I’m okay with all of this is that I just don’t care if a machine is looking at my pictures. It would be different if, you know - an actual person - was ogling at my candid selfies. I’m sure a lot of people feel the same way. And so millions of us will feed data into the machine. That machine will only get smarter and smarter. So, what happens when that machine knows us better than we know ourselves? An interesting question, but I’m having too much fun organizing my pictures to think too much about it right now.



  « Previous: Next: »