The voice assistant seemed so cool and made so much sense. What happened?
If you follow Internet of Shit, you already know. This Bloomberg story about Alexa adoption suggests even the biggest players in voice-enabled hardware are struggling to find their why.
We have several Echos in our house, all of which are used for extremely banal reasons that are just easier than connecting dumb speakers by bluetooth for the most part, or pressing buttons on an oven timer. Of course, they were envisioned as transformative technology, not egg timers.
Bloomberg doesn’t go deep on this, but points to the overall onboarding experience as being where users seem to check out, if not shortly thereafter. Some of the complaints are familiar: in an effort to get you to do more with the device, it starts asking you if you’d like to try new skills, few of which have any relevance to what you’ve been doing with it.
Dieter Bohn wrote how we’re still getting voice assistants wrong for the Verge in 2019 and if you search the title of this post, you’ll find no shortage of commentary. Heck, Android Authority asked their readers if they used them and the overwhelming majority said no. To be clear, this isn’t that they struggle with natural language processing — they absolutely do — it’s also that they don’t seem to connect to any patterns in usage and then develop intelligence built around it. Isn’t this what AI and machine learning is all about to the layperson?
It’s not just Alexa; it’s every voice assistant I’ve ever used. If you’ve found any Alexa skills or routines useful, share them below.