Here’s what’s next for Alexa

The Big Story

What's next for Alexa and voice interaction

Amazon held its annual Alexa Live developer event yesterday, which the company used to announce approximately a bajillion new features and developer tools. Honestly, it was a lot — too much to dive into here. However, it's worth highlighting a few key themes coming out of the event:

• Voice is becoming a true developer ecosystem. Amazon now has more than 900,000 registered Alexa developers, who collectively have built over 130,000 Alexa skills. And those skills are starting to show up in more and more places: "We now have hundreds of physical products" with Alexa built in, Alexa Voice Service & Alexa Skills VP Aaron Rubenson told me this week.
• It's getting easier to make money with ambient computing, as Amazon is now allowing developers to charge upfront for their skills, much like you'd pay for an app on your phone. Previously, developers were only able to sell extra content from within free skills. "We're very focused on helping developers drive business growth," Rubenson said, adding that in-skill purchasing more than doubled year-over-year.
• Discoverability will be a key issue going forward. Around 130,000 skills is a lot to sift through, which is why Amazon is adding some discovery features. These include personalized skill suggestions, and ways for Alexa to suggest a curated list of skills for common queries.
• Alexa is finding its way into legacy media. Amazon is looking to enable a bunch of interactive voice experiences for media companies, starting with the ability to request songs from radio DJs via Alexa skills. "The initial launch is really focused on radio, podcast and music providers," Rubenson said.
• Alexa is getting more playful. Asynchronous multiplayer games and experiences are coming to Alexa devices, thanks to a new Shared Activities API that will officially be released later this year.

In addition to announcing all these new features, Amazon also used Wednesday's event to put a stake in the ground and signal the world that we will see a lot more of Alexa — even if we won't always recognize her.

• Verizon's upcoming smart display was officially announced at the event. The display makes use of a white-labeled version of Alexa that offers access to Verizon-specific features. People joked about the device's "Hey Verizon" wake phrase when Protocol first reported on it a month ago, but Rubenson argued that it made sense for well-known brands to bring their brand expertise to ambient computing. "Customers already have well-formed perspectives on the superpowers of a given brand," he said. "I think that customers will love this device."
• Amazon is doubling down on voice interoperability. The company released a new toolkit for consumer electronics companies looking to run multiple agents on their devices. This will put a lot of pressure on Google to rethink its resistance to a multi-assistant world.

During our conversation, Rubenson also admitted that there are still challenges ahead, some of which Amazon addressed in two new white papers released yesterday. What happens, for instance, when one person asks Alexa to set a timer, and then a family member walks in and asks another assistant to stop it?

Rubenson's answer: The second assistant should be able to stop the timer. "That doesn't mean that every agent has knowledge [of] what the timer is about," he said. Voice assistants should be able to cooperate, but limit and be transparent about data sharing, Rubenson said, adding: "We are just stepping very slowly into these use cases as they are getting more complicated."

Overheard

"Being compared to Google and Facebook is Netflix's worst nightmare." —Bloomberg journalist Lucas Shaw, pointing out that Netflix would prefer to avoid the kind of regulatory scrutiny Big Tech has been facing.

"It's a multi-year effort. We're going to start relatively small. We'll learn, we'll grow, we'll refocus our investment." —Netflix Chief Product Officer Greg Peters on the company's entry into gaming.

A MESSAGE FROM ENVOY

When it comes to hybrid work, let's be honest: People don't know what to think. Are two or three days a week in the office the sweet spot? Are five days a dealbreaker? We repeat what we hear others say or go with our gut, but how do any of us really know?

Learn more

Watch Out

India's BYJU's buys reading app Epic for $500 million

Indian educational software giant BYJU's has acquired the Bay Area-based reading app maker Epic (not to be confused with Epic Games) in a stock and cash deal valued at $500 million. Epic will continue to operate as a standalone service, and its two co-founders, Suren Markosian and Kevin Donahue, as well as their staff of 160 will all join BYJU's.

Never heard of Epic? Then you probably don't have any primary school-age children.

• Epic has been growing at a rapid pace since the start of the pandemic, Markosian and Donahue told me during an interview this week. The company's apps have been used by 50 million children to read over a billion books during the past year.
• Much of that growth has to do with Epic's freemium model. Epic is making its service available for free to schools, and is charging parents for access to additional titles.
• Now, Epic wants to go overseas. The company's business has been primarily focused on North America, and Markosian said that one of the goals of the acquisition was to get help expanding internationally. That's very much a two-way street for BYJU's, which is big in India but virtually unknown in the U.S. "We can make it a global brand, and a global company," Markosian told me.

That aligns well with the plans of BYJU's, another fast-growing company you may have never heard of. Founded a decade ago in India, BYJU's has become an educational tech powerhouse whose apps are being used by 100 million students.

And there's an AR angle, too: Two years ago, BYJU's spent $120 million on OSMO, the visual computing startup that had figured out how to use an iPad's front-facing camera to expand the play time beyond the screen. OSMO and BYJU's have since been busy working on their own learning-centric tablets and smart displays, and I wouldn't be too surprised if those devices eventually run Epic apps as well.

A version of this story first appeared on Protocol.com.

Fast Forward

• On Protocol: Boomy's AI has created more than 3.2 million songs. Some of the service's biggest fans are TikTok creators looking for some extra income.
• Facebook's Quest VR headset may be getting a storage bump. Leaks suggest that the company is doubling the Quest's base model's space for apps.
• Android TV is getting Google's Watchlist.Step by step, Google is making legacy Android TV devices look more like Google TV.
• Kids adopt British accents because they're binging "Peppa Pig."This is such a great story about lockdowns, Peppa and the power of streaming.
• On Protocol: AR and VR aren't on the podium in Tokyo. Producing immersive experiences for the Olympics has been challenging during COVID-19.
• CNN will launch a CNN+ subscription service next year. No word on pricing yet, but the news network apparently isn't planning an ad-supported version at launch.
• Startups bet on short-form snippets as the future of audio.Here we go again. Doesn't anyone remember Odeo?

Auf Wiedersehen

Years ago, Shazam's developers received emails with an interesting feature request: In addition to identifying songs with their phone, people also wanted the app to tell them which birds were chirping in their neighborhoods. "We wanted to, but couldn't invest/distract to pull off this massive feat," recalled former Shazam VP Katie McMahon. The good news: The Cornell Lab of Ornithology just added a "Shazam for birds" feature to its own bird-watching app. The app immediately got a wholehearted endorsement from McMahon, who knows a thing or two about audio recognition. "So impressive," she tweeted. But don't take her word for it: The app is free, so install it, have it download a few necessary extra files and you're ready to get a lot more familiar with those winged little friends in your backyard. Happy birding!