Amazon wants Alexa to make friends: The company announced a new toolkit at its Alexa Live summit Wednesday that will make it easier for companies to build devices that run multiple voice assistants at the same time. Amazon is also doubling down on white-labeling Alexa, and officially announced a Verizon-made smart display with a custom-branded assistant that Protocol first reported last month.
In addition to its Verizon partnership, Amazon also teamed up with LG and Samsung to build devices with multiple voice assistants. Future LG TVs will run both LG's own assistants and Alexa at the same time, and Samsung's smart refrigerators will offer access to both Bixby and Alexa. Taken together, the announcements put further pressure on Google to change its stance on voice assistant interoperability — an issue that has recently received attention from senators looking to rein in Big Tech.
"We envision a future where the world will evolve to have multiple ambient assistants," Alexa Voice Service & Alexa Skills VP Aaron Rubenson told Protocol ahead of Wednesday's event. People should be able to pick whatever assistant they prefer for any given task, simply by invoking different words, Rubenson said. "We think it's critical that customers have choice and flexibility," he said. "Each will have their own strengths and capabilities."
Amazon's efforts to build this multi-assistant future include its spearheading of the Voice Interoperability Initiative, an industry consortium that now has close to 90 members, as well as practical support for partners looking to build multi-assistant devices. In addition to LG and Samsung devices, these also include Vodafone, Deutsche Telekom and Facebook, the latter of which integrated both its own assistant and Alexa into its Portal smart displays.
To simplify this kind of integration, the now-announced toolkit includes both middleware components and developer guidelines. "The toolkit was informed by the work that we have done with other partners that have built implementations from scratch," Rubenson said.
On the one hand, this is very much the kind of in-the-weeds technical work that one would expect to be announced at a developer event. On the other hand, it's a highly political broadside against Google and its stance on voice assistant interoperability. The search giant has long prevented manufacturers from building devices that simultaneously run the Google Assistant and other voice assistants. During a recent Senate hearing, a Google representative argued that this was due to unresolved technical challenges.
By simplifying the development of multi-agent devices, Amazon is very much undermining that argument. Rubenson told Protocol that he hoped Amazon's stance on interoperability would be bolstered by additional devices running multiple assistants. "We know it's possible today," he said.
Sonos Chief Legal Officer Eddie Lazarus made a very similar argument during last month's Senate hearing. When Google's senior public policy director, Wilson White, told senators that technical issues would prevent companies from implementing the Google Assistant alongside other assistants, Lazarus said his company had developed a working solution, but was prevented by Google's licensing terms from using it.
This prompted an unusual exchange, with Wilson signaling that Google may be open to revisit its stance, and that he'd personally love to see a demo of Sonos' tech. "This will evolve," Wilson said during the hearing. "We will get to a place where we are bringing more innovation to consumers."
"We were encouraged [by] the dialogue at the hearing," Rubenson told Protocol this week, adding: "We would welcome Google's participation in the Voice Interoperability Initiative."