Smart speakers were the hottest thing in home automation five years ago. Almost everyone wanted one even if a lot of people still didn’t buy one. These days however, the smart speaker seems to have stagnated. Year-on-year sales have definitely slowed since COVID. Will the technology eventually die, or will it be resurrected by the latest thing to capture America’s attention: artificial intelligence (AI)?
We have known about AI for quite some time. Most of us have only given it a fleeting thought every now and again. But the emergence of ChatGPT has brought AI to the forefront. ChatGPT is to modern AI what the smart speaker was to home automation back in 2018. Marrying the two seems perfect, doesn’t it?
That is exactly what’s happening. At least one AI company, Josh.ai, is working on a smart speaker that utilizes artificial intelligence utilizing OpenAI’s now-famous ChatGPT. The implications of their project are paradigm shifting.
Why the Smart Speaker Has Stalled
Data demonstrates that smart speaker growth was explosive right up until the COVID pandemic hit. Then it stalled. Here are recent statistics showing the number of U.S. residents who reported owning a smart speaker:
- 2018 – 47.3 million
- 2019 – 66.4 million (42% year-on-year growth)
- 2020 – 87.7 million (32% year-on-year growth)
- 2021 – 90.7 million (3.4% year-on-year growth)
- 2022 – 94.5 million (4.6% year-on-year growth).
It is reasonable to blame the double-digit crash in growth between 2020 and 2021 on economic factors related to COVID. But those same factors were not in play the following year. So why has the smart speaker stalled? My guess is limited functionality.
A smart speaker is a novelty in the early days of ownership. That makes it popular. But as a standalone device, it doesn’t do much to enhance the lives of people who aren’t already interested in home automation. A smart speaker’s limited functionality becomes its own worst enemy. But throw in artificial intelligence and you have a whole new ball game.
Multiple Things Change with AI
Right now, smart speakers offer limited function as standalone devices. They can be combined with home automation systems for the purposes of voice control and creating home automation routines. Vivint Smart Home is just one home automation provider whose systems are compatible with both Google and Amazon smart speakers.
All of this is well and good except for the fact that smart speakers are trained to recognize static commands. The limits of natural language processing (NLP) prevent them from understanding and responding to nebulous commands that don’t follow a static formula. That is where AI comes in.
Effective AI still needs effective NLP to work in a smart speaker scenario. But AI can actually improve NLP by eliminating the need to stick with a set of preselected commands. AI can analyze what a smart speaker ‘heard’ and compare it with all sorts of possibilities to make a decent guess at user intent. Then it can scour the internet for information pertaining to that intent.
Greater Dependence on the IoT
If there is any downside to combining AI with smart speakers, it is the fact that the resulting technologies will require even more dependence on the internet of things (IoT). For people who are already internet averse, turning more of their lives over to the IoT will not be optimal. But for everyone else, it’s a small price to pay for AI-powered smart speakers.
Time will tell if AI can breathe new life into the smart speaker. I am betting it can and will. Combining the two technologies seems like a marriage made in heaven.