Smart Speakers Are Planned Obsolescence in Real Time

The Amazon Echo speaker (and Google Hub, etc…) are an interesting window into how tech companies push technology into our lives. They are in the form factor not meant to last, but are simultaneously pushing a technology onto us that is meant to be the next phase of computing, using a largely voice-based OS. And, now that smart speaker functionality is starting to get baked into a lot of other devices like Air Pods, FireTV remotes, and a million other places, the smart speakers themselves will start to disappearing.

It’s rare to see a whole product category that can be described as “planned obsolescence”, and it looks like the “obsolete” part of smart speakers is starting to kick in. I, like most people, use my Echo for fairly trivial things like playing music and setting timers. And while there was a lot of optimistic talk early on, it seems like new use cases for smart speakers are not appearing as we thought they would. Simultaneously, the functionality that does exist is getting rapidly rolled up with all of our other devices — I can now change the music in my house using voice commands from my phone, my tv remote, my headphones, and my computer just as well as I can from one of my physical Echo devices. Not surprisingly, I do not use these Echo devices themselves as much as I used to, and I would bet that I’m not the only one.

I get push back on the idea that this product category actually is “planned obsolescence”. After all, why would you release a tech product you knew was going to be superfluous a few years down the line? My response is to point to the (often) $25 price tag for the Echo Dot. I have no idea what they cost to manufacture these devices, but they sure as hell aren’t making much of a profit at $25, and while the higher end speakers go up to $200 or $250, this is still quite inexpensive by tech standards. Of course, I think these companies would love for these smart speakers to be insanely successful, but I also think this was the plan B. Plan A was, and remains, to get us used to voice computing while simultaneously improving their ability to successfully perform voice computing on the backend. Natural language processing takes a lot of real language data, and teaching us humans to learn a new form of computing using our voices takes a lot of time. Smart speakers have been merely been the training ground for both of these things to happen.