Google’s Brick Factory

Back in 2013 I debuted my How Personal Is Your Personal Cloud presentation to the San Francisco Personal Cloud meetup. My objections to the prevailing #IoT architecture are many and among them is the notion that it’s crazy to build an entire ecosystem of personal devices around an architecture that only works with an Internet connection to the vendor.

I tried shopping this idea around IBM where I was working at the time, including giving the same presentation internally and speaking with some of the Distinguished Engineers and STSM’s in the IoT space. Most weren’t interested. A few allowed that the architecture I described was probably better for the individual customer but not strategic to IBM whose customers are large enterprises.

I remained so convinced that this vendor-owns-your-data architecture is bad that it played a large part in my decision to leave IBM. I was unable in good conscience to look a customer in the eye and say with a straight face that I had the best solution for delivering device data to the enterprise. That part is actually true but it assumes as a forgone conclusion that getting that personal data, whatever it is, to the enterprise is the desired outcome and the ultimate goal.

In the following three short years we’ve seen several examples of just how bad this problem is. In his essay The time that Tony Fadell sold me a container of hummus Arlo Gilbert describes how his highly integrated home, built around the Revolv hub which was bought up by Google, is about to be bricked. The problem is that Revolv wasn’t too successful and so not many people are invested enough to care. But as I point out in How Personal… and Arlo points out in his essay, the fact that its Google (now Alphabet) shutting down the service should make this bigger news.

1. How many devices do you have?
2. Of those, how many are bricked without an Internet connection?
3. Of those, how many are worthless if the vendor stops service or goes out of business?

If the answers to #2 and #3 are not significantly lower than the answer to the previous question, you have a significant exposure.

At what point do we as an aggregate market begin to care whether our devices work stand-alone? Or solely within our own personal network? When do we claim our personal data as OUR personal data and reject the notion that every hardware and software vendor has a natural entitlement to it?

Most importantly, will our “smart” devices ever not be so incredibly stupid in their architecture if we don’t demand a change? Probably not. If it wasn’t painfully obvious before, thanks to Google we can now see the hummus on the wall and it isn’t looking good. So what are we waiting for? Time to stop buying IoT devices that are not much more than vendor data portals into your personal private space. Time to vote with our money for architectures that we control and we own.