Amazon’s Alexa is leading the AI assistant pack. Echo devices are dominating smart speaker sales, and that was before Amazon brought the devices to more than 80 nations around the world. To defend its crown, Amazon moved fast this year to outpace competitors like Google Assistant and Microsoft’s Cortana. Apple’s delayed HomePod is due out next year, while Samsung and Facebook are also reportedly planning to debut smart speakers. It can be challenging to keep up with all the features Alexa has added to stay ahead of some of the largest tech companies on the planet, so here’s a rundown of everything Alexa learned to do this year.
If you’re the new owner of a smart speaker like a Sonos One or Echo Dot, see this story about 10 things to try with your Alexa-powered device.
A more expressive voice
Earlier this year, Speech Synthesis Markup Language (SSML) tags were introduced to give Alexa a more expressive voice. More natural voice features were made available to the makers of Alexa skills in the U.S., the U.K., and Germany.
In the United Kingdom, SSML can be used to train an Alexa skill to say “blimey” or “Bob’s your uncle,” while skills can be trained to say “bada bing” or “boom” in the United States. SSML can also be used to make Alexa whisper or change her tone of voice, along with many other effects. Siri and Google Assistant also gained more natural or expressive voices this year.
This is fun, especially if playing music is the main reason you like having Alexa around.
Earlier this month, ahead of the release of the Echo Spot, Alexa learned how to create musical alarms. That means you can choose from millions of songs to wake up to every day.
Alarms can be set based on artist, genre, or song name, or even based on lyrics from a favorite song whose name you’ve forgotten.
If you live in a house where the rights to DJ Alexa music volley back and forth, good luck to you.
Alexa goes beyond the Echo
At the start of 2017, Alexa was basically only available in the Echo, Echo Dot, and Echo Tap, but this year we saw the introduction of the second-generation Echo and the Echo Plus with a Zigbee smart home hub.
Also released by Amazon this year: the Echo Look, which uses computer vision to be your fashion stylist; the Echo Show, which introduced Alexa’s first visual interface; the Echo Spot, to extend that visual interface to your bedside and replace your alarm clock; and the Echo Wand, a remote-like device with a barcode scanner and Alexa inside that sticks to your fridge.
And Alexa isn’t just in devices made by Amazon anymore. This year, Alexa was made available in speakers from Sonos, Harman Kardon, Lenovo, and other brands.
Create reminders and add events to more calendars
In its rush to be your everything, Alexa launched without some features you may consider rather basic. Among them: reminders. How this feature remained unavailable for so long is beyond me, but in June, Alexa finally learned how to set reminders.
Make phone calls and send messages
Alexa is now able to make free phone calls in Canada and the United States. You can also use Alexa to send messages to other users of the Alexa app or Echo devices. The back-and-forth between Alexa and its main competitor, Google Assistant, has been an undertone this entire year, and the introduction of phone call and messaging features is no different.
Whereas Google Assistant is able to send SMS messages, Alexa still only sends messages via the Alexa app, a feature first made available this spring.
Alexa as home intercom
This summer, Alexa learned how to broadcast voices to Alexa devices and act like a home intercom system. Since the average Alexa device owner purchases more than one device to make voice control available throughout the home, this makes a lot of sense.
Because it is made available through Drop In, Alexa’s voice and video call feature, you don’t even need to be home to yell at your kid or ask whoever’s in the kitchen to check whether you have milk.
25,000+ voice apps
At the end of 2016, Alexa had roughly 5,000 skills. In the past year, that number has quintupled to more than 25,000. This far exceeds the third-party offerings from competitors like Google Assistant or Cortana, who only have hundreds of voice apps.
New skills added in roughly the past year include This Day in History from the History Channel, the Twitch Alexa skill for streaming video gameplay, and series made especially for kids or for Echo Buttons.
A great addition in the past year is the ability to discover Alexa skills without saying its exact name. The need to remember the name of a skill is a limitation of the voice-first interface, but now you can say things like “I need a ride” to hear recommendations like the Lyft skill.
Whereas last year you could ask for a joke, now you can ask for a joke about dogs or trees or whatever’s on your mind. If you play a popular game skill like The Magic Door, you may now receive a recommendation to play another popular game.
Fleshing out discovery and recommendation features may give developers more incentive to make Alexa skills, and it could teach Alexa device owners that they no longer have to remember the name of their go-to voice apps.
Double Jeopardy! was the first Alexa skill to be offered with a paid subscription. Next year, more paid subscription or one-time payment Alexa skills will be made available, starting with Teen Jeopardy!, Sports Jeopardy!, Match Game, Heads Up, and History’s Ultimate History Quiz.
Alexa, your personal gaming assistant
In addition to the expansion of Alexa recommendations when you play a game skill or say “Alexa, let’s play a game,” for much of 2017 the Alexa team at Amazon has worked to improve the assistant’s game-playing chops.
This past spring, Amazon began to pay skills developers based on the level of engagement their skills are able to create, starting with game makers.
Last month, Alexa began to act as a personal assistant for Destiny 2, one of the most popular games of 2017, and last week saw the debut of Echo Buttons, a simple, first-of-its-kind device made especially for multiplayer gameplay.
Next year, Alexa will begin to act as the host for a set of board games called Voice Originals, starting with When in Rome, which was created by Sensible Object, a graduate of the inaugural Alexa Accelerator. Next year, the Gadgets API and SDK will also extend game skills that interact with Echo Buttons and other hardware made by gaming companies.
Alexa perks for Amazon Prime members
This year, Amazon began to extend a steady stream of exclusive perks to people with both an Echo device and Amazon Prime. In March, two-hour deliveries were made available through Prime Now for things like groceries, and even alcohol in select cities.
In July, Amazon made it possible to get a $99 Amazon Prime annual membership with an Echo device.
This fall, the Double Jeopardy! skill was made available free of charge to Echo device owners with a Prime membership.
Last year, Amazon CEO Jeff Bezos posited that Alexa could become one of Amazon’s four pillar offerings, alongside the company’s marketplace, AWS, and Amazon Prime. Alexa is clearly a big part of Amazon’s roadmap, but to what extent Amazon unites Prime and Alexa in the years ahead remains anyone’s guess.
Unique voice recognition
Alexa can now recognize up to 10 unique voices in a household, beating out Google Assistant’s ability to recognize up to six voices. Unique voice recognition gives Alexa the power to make sure kids don’t add items to the shopping list or ensure only your calendar events are mentioned if you ask Alexa about your schedule.
Alexa skills developers will be allowed access to unique voice identification provided by Amazon next year, giving them the ability to remember past interactions and deliver more personalized experiences.
Unique voice recognition is also being used to help Amazon recognize a user’s emotional state or mood. This effort will begin with understanding when a user becomes frustrated so Amazon can determine whether Alexa responded to a question or command in the right way. No timeline or details about recognition of emotions beyond frustration have been provided, but emotion recognition could eventually help Alexa respond to you more empathetically, or learn your emotional baseline to recognize when you’re lonely or having a bad day.
Smarter Alexa skills
In addition to the introduction of a visual interface with elements like cards and videos, this year developers were given the ability to access user location so, for example, a restaurant or other business can recognize where you are and serve up customized results or make deliveries.
Better smart home voice control
In 2017, Alexa gained the ability to do things like change the color of a room with smart lightbulbs, so if you’re feeling orange, just say so. Alexa also learned to control smart home cameras and share security camera footage through an Echo Show or a television with a Fire TV stick.
Other important smart home milestones reached this year include the ability to group Echo devices together for music throughout the home, and hands-free Fire TV control with Alexa.
In what could turn out to be a momentous event in the history of AI assistants, this summer Seattle-area tech giants Microsoft and Amazon agreed to pair Alexa and Cortana across each other’s devices so that Cortana will be available through tens of millions of Echo speakers and Alexa will be available through more than 500 million Windows 10 PCs.
This appears to be one of the biggest partnerships of its kind to date. When the partnership was announced, Amazon CEO Jeff Bezos talked about assistants growing into areas of specialization, but that doesn’t match up with rhetoric heard from the Alexa team at Amazon, who say they want the AI assistant everywhere. Indeed, there are a number of advantages and disadvantages to a Cortana-Alexa partnership.
For example, last month Alexa for Business launched at AWS re:Invent, an initiative that is sure to compete directly with Cortana’s apparent focus on productivity.
In 2018, we should get some answers to how the Alexa-Cortana partnership will work in practice and the kinds of limitations these competitors put in areas like the workplace, where both want to make inroads.
Routines for command customization and scheduled actions
Routines allow you to create custom voice commands and trigger multiple actions with the utterance of a single word or phrase. Today, you can do things like make a routine for when you wake up in the morning that turns on your lights, gives you a weather report, tells you about traffic for your commute, and reads you the news. Routines can also schedule tasks so you can do things like play the news at a specific time every weekend.
You can also use this skill for parlor tricks, so when you say “Alexa, get turnt up,” she responds by saying “Let’s party.” Routines that help you get stuff done, like reading your to-do list or your calendar events, seem like a natural next step.
Japanese and new English dialects
Last year, Alexa and Echo devices expanded to users in the United Kingdom and Germany, trickling into the assistant’s two first countries outside the U.S. with things like the Flash Briefings API and Alexa Voice Service access to extend Alexa into devices beyond Echo smart speakers.
In late 2017, Alexa learned to speak even more languages, with support extended to English speakers in Canada and India and to Japanese speakers in Japan. Next year, Alexa will be available to English speakers in Australia. For now, international versions of Echo devices rely on these language settings, though the expansion to more than a dozen Spanish-speaking countries alludes to the fact that Alexa will soon be able to habla español.