Alexa, Find A New Name and Mission

My daughter had a classmate whose mother’s name was Siri. On hearing that, my first thought was that I bet her day to day experience took a real turn a few years ago. It turns out, this is definitely a thing. There aren’t all that many Siri’s out there in the US, but there are a lot of women named (or at least who go by the name) Alexa and they are mad. Any time I’ve mentioned something like this on Twitter I get visits from Twitter accounts dedicated to anger on this subject, and a movement. And honestly they have a point.

An AI is chipsets and programming; it has no attachment to any name and can activate by whatever word its programmers tell it to. Giving @alexa99 a real name and referring to it as “she” is causing serious harm to people, and it’s completely unnecessary for the AI’s functioning. https://t.co/oCv4idAIey

— I Am Alexa Alliance (@AlexaAlliance) November 29, 2022

Samsung and Microsoft, with Bixby and Cortana respectively, seem to have the right idea. Those are namey and they sound like names but at minimum cause disruption for fewer people.

On the other side of the spectrum is Google’s product, which goes by Google Assistant and has you say “Hey Google” on the street like a crazy person.

What’s not clear to me is why they need a singular name at all. As an exercise in branding “Siri” or “Alexa” may make sense (though again Bixby or Cortana is probably better) but I guess I always figured that they (and especially Google) would eventually let you name your own assistant. People would likely gravitate towards uncommon names to avoid confusion and it would be spreading the pain around a little bit (and honestly wouldn’t be too unlike when we’re out and about and hear someone calling for someone else by our same name). Except people named Alfred, who are probably used to stuff like all that.

I use a third-party app on my phone to alert Alexa, and that app lets you at least add a word to the name. Which might be better for all the Alexas out there if people use it with thought. I gave my Alexa the last name McIntyre, so I say “Alexa McIntyre” and it wakes up.

Unfortunately, I have to do this because Amazon doesn’t support voice waking on most cell devices and are discontinuing support on the rest next April. This is part of a broader scaling back of voice assistants. Android is laying off Alexa staff, Microsoft killed the Cortana app for Android earlier this year, and apparently Google did the same a little while back. I’ll probably never be able top name my assistant, alas.

The main problem with the digital assistants is that they are failing to actually generate revenue. It has been decided that this is a problem. I think there are problems in mission and execution that have plagued all of them (except Siri, I suppose). They misunderstood the question of “What are they for?” Amazon actually expected people to order stuff with their voice and I don’t think it was ever going to be easier to buy something conversationally instead of looking at a bunch of products and prices in grid form on a computer or device screen. The closest I get to that is telling my assistant to add something to my “Amazon List” or “Walmart List” which I will later search. But neither Google nor Amazon (the two I use) do it well and in fact do worse than they did a couple of years ago. Alexa and Google Assistant are getting dumber (or, more precisely, losing support for various activities).

If these things are to have a purpose it should be to keep us in their ecosystem. The number one way that Alexa could have me using Amazon Music is by allowing me to say “Hey Alexa, play Phil Pritchett” and have it start playing Phil Pritchett. But now to use “Hands Free” Alexa I have top actually open the app first then talk to it, which is not hands free. The closest I can get to actual hands-free is saying “Hey Google, open Alexa” followed by “Alexa, play Phil Pritchett”. This is better than nothing, but not useful. When Google Assistant stopped being able to do this (when I told it to play an artist while driving – the number one time I want to use hands-free – it said that I couldn’t do it on the device. It was a good opportunity to get me to switch over to Android Music and they blue it by failing to do what Some App by Some Guy in the Google Play store accomplished (automatic waking when I say “Alexa McIntyre”)¹

So I call that a failure of mission. What should the voice assistant be for? Privileging your own ecosystem. The core of that would not have required nearly as much effort as what they were trying to do. It just needs to play Phil Pritchett when I tell it to play Phil Pritchett, add stuff to the task list when I tell it to add stuff to the task list, and open up a navigation app with the location when I tell it I want to go somewhere. From there you can branch out, but that should have been the core competency and neither were super competent at doing it.

That leads to the failure in execution, which is that like a lot of AI it seems to think in terms of word clouds and then try to piece together what is being asked of it. Trying to figure out hard syntax became weirdly difficult at times. I gave up on being able to play the Lumineer song Cleopatra because no matter what words I used it would start playing the album. Saying “Hey Google, play THE SONG Cleopatra by The Lumineers” should have gotten the job done. Or if you wanted to use the words in a different order “Play The Lumineers’s song Cleopatria” that’s fine. Either way, just give us a way to look up the syntax if we’re having trouble. Instead they often try so hard to figure it out *for* us that it gets some basic stuff wrong. Microsoft Word auto-formatting writ large.

I had always assumed that they were going to fix this eventually, but maybe not.

1. Unfortunately, the app in question, Ultimate Alexa, can do some things with Alexa but playing music isn’t one of them. I still use it to add stuff to lists.