Are voice interfaces really the future?

Patrick Sanwikarja
3 min readMay 30, 2022

--

Google Home Mini, the little speaker that you can talk to, has been in our home for four years now. Even before it became available in The Netherlands, I bought one on a trip to London. A device that I can talk to and connect to my smart lights? I just had to have one. It’s the future!

When I could officially buy the devices in The Netherlands as well, I thought: what the heck, let’s put a Google Home in every room of the house! So now, wherever in the house I am, I can talk to Google. Because voice interfaces would be ‘the future’.

Or so I thought.

Because it appears to me that the quality of the Google Assistant hasn’t improved — in my experience it has actually gone down since I bought that first one. It has more trouble of understanding what I say. Sometimes the wrong speaker reacts. And sometimes it doesn’t hear me at all, even though I’m standing right next to it. When I ask it to play music on Spotify, it just stops playing halfway the first song.

Don’t get me wrong — I still use Google Home all the time, every day. It’s just that I’m underwhelmed with how ‘smart’ it is.

I remember being quite impressed in the beginning, with what I could ask Google. Not just to turn our Philips Hue lights on or off, but also everyday questions like what the weather is, setting timers and reminders, asking for trivia. And I could install apps on it, like the app from Albert Heijn, my supermarket, which I could use to add items to my shopping list. That was a real killer feature. But Albert Heijn discontinued that app already 2 years ago.

So much for voice interfaces being the future.

Or is there hope after all? Because today, for the first time in a long, long time, Google again impressed me.

We were having dinner and were discussing what my girlfriend and I should be wearing to her cousin’s wedding this weekend. And if perhaps I should get a haircut, and if so, what hairstyle I should get.

“Why don’t you get a mohawk again”, my girlfriend jokingly suggested.

“What? Did daddy once have a mohawk?” my daughter asked surprised.

“Absolutely”, I said. “But that was a long time ago. I’ll show you a picture later.”

But my daughter couldn’t wait. “Please can you look up the photo now? Please?”

I said it would have to wait until after dinner. And my phone was in the charger.

“But I want to see it!”

Then I thought: let’s try and ask Google.

Near the dinner table we have a Google Home Hub, which has a small display that normally acts like a dynamic photo frame but can display information related to what you ask Google. I didn’t expect Google to understand my question, but asking is free, so I said:

“Hey Google, show pictures of me with a mohawk”

And sure enough, after a second Google replied:

“Here are pictures of you with a hairstyle that matches a mohawk”

The display showed a picture of me with a mohawk.

“Holy shit, it worked!” I reacted.

“Daddy you can’t say that!” my daughter corrected me.

“Oh sorry. But look, there’s a picture of my mohawk. Cool, right?”

Google had recognized my voice, understood my question, and searched my Google Photos libary for my face and a mohawk on my head. It only found one photo, even though I’m pretty sure I have multiple photos of my mohawk-days. But still, it worked! Pretty damn cool.

I don’t really understand how voice assistants evolve. You would expect that they gradually improve the more you use them and the longer Google can work on its algorithm. But perhaps their evolution is more like natural evolution: it’s not a gradual process, but it happens in leaps and bounds.

--

--