At its Worldwide Developers Conference earlier in June, Apple made a flurry of announcements: new Macs, iPad Pro 10.5-inch, iOS 11 and WatchOS 4. The biggest announcement was Apple’s new Siri speaker, HomePod. HomePod is poised to compete for home field advantage with Amazon Echo and Google Home. Previous to this announcement, Google and Amazon had already begun a battle to control our abodes.

The HomePod’s features are comparable to those available for Home and Echo, so at this point the forces driving the decision seem to be platform loyalty and pace of adoption. Were Apple aficionados waiting on an Apple device? What about iOS users who wanted to get into the smart speaker market early? What about the anything-Android-Google-groupies? At this point it’s too early to say who will emerge as the leader in smart speakers. This hasn’t stopped people from trying to determine which AI earns an “A.” Fortune recently reported on a 3,000 question quiz on which Home outperformed Alexa.


So far, all of the smart speaker speaking points have been based on the single user, single household model. Settings, preferences, accounts and ads have all been tied to individual users. This makes sense since this has been the standard operating model for the decade since the smartphone revolution began. What if, ten years on, the accounts and answers were tied to spaces instead of users. What if the smartest things in the room were not the speaker but the room–its questions, its history, its combined power to query?

What if, ten years on, the accounts and answers were tied to spaces instead of users. What if the smartest things in the room were not the speaker but the room–its questions, its history, its combined power to query?

Now there are two questions that I am asking myself as a language teacher. The first question is why it has taken so long for any of these three companies to come to market with a speaker that understands languages other than English? Engadget reported that Google Home and Assistant can speak Canadian French now, ahead of it launch in Canada on June 26th. According to announcements at the I/O event, French, German, Brazilian-Portuguese and Japanese will come online by summer’s end. By the end of the year, the Assistant will also be able to speak Italian, Spanish and Korean. How will this push allow Google’s AI Assistant to dominate in new and existing markets internationally? Closer to home, how will voice enabled AI inform and inspire language learners in the United States?

This line of inquiry about the role of voice enabled AI in all these languages–including English–leads me to my second question: How can smart speakers empower groups of learners to ask interesting questions and begin to distinguish between answers AI can provide and those that only a thinking being–peer or professional can provide? Can a classroom benefit from this resource or would the volume and variety of questions break things down? Even the the speaker could handle the “volume and variety of questions,” how could it account for the volume in the classroom space?

“Hey, Siri…what role will smart speakers play in schools next year?”

Let’s imagine a few scenarios: a 3rd grade classroom, a high school Spanish class and a shared middle school learning space (e.g. library or lab). I have chosen these for two reasons: a) because they represent a variety of K-12 learning environments and b) they are learning environments I inhabit most days.

3rd Grade Classroom – “What is pee?” Yeah, I’ll just put that out there. This would undoubtedly happen and any AI would have to build in some snappy responses, equal part sass and science. Once that question is answered, the remaining questions at this level would have to be when do you allow for AI? Why? and with what consistency? For instance, there may be points where spelling, multiplication and states and capitals were strictly off limits. Part of learning at this level is memorization. However what about review questions? What about exploratory questions–perhaps not “What if…?” at this level, but “Why…?” Would there be a benefit to having the speaker in the space, either because it could come pre-programmed with grade/age appropriate content or because it could learn from the current students? Could the speaker distinguish among voices and cater responses accordingly? Would doing this create a serious privacy problem? Could the data be cleared and cached in a responsible way so that the nature of the questions and the real “nuggets” were saved but the others were deleted? Additionally, AI could be consistently integrated as a dependable assistive technology for all students. If most of these questions could be answered to a satisfactory degree, we could be looking at an AI TA. (Call her “Aitana.” You heard it here first.)

High School Spanish Classroom“¿Cómo se dice ‘pee’?” Some things  don’t change. Seriously, most questions in this environment would be vocabulary based. This is understandable and perfectly appropriate. Beyond vocabulary, there may be some really interesting applications. Teachers can work with students on ways to ask questions–after all the response is a function of the information available and the correct “code” of the questions. “¿Qué tiempo hace en Bogotá?” therefore becomes a question about the weather in Bogotá as well as an exercise in syntax and pronunciation. Ditto for all questions about when events happened and who the primary players were. This could be a real boon for language classes. One of the main issues is that a “smart” speaker in this environment would need to be able to interpret a question from a novice speaker. What margin for error can programmers build in? Could this be a function teachers could toggle? What sort of reverse engineering would would be necessary to account for interference from a student’s native language? These are questions that would need to be answered, but there is untold potential in language study. Add some songs, soccer talk and jokes to this mix and the technology could have meaningful applications if the technology matures.

Shared Middle School Learning Space – “‘Sup, Siri?” The chaos in this environment could mean that an AI-enabled speaker is either the greatest idea or the grandest fail. Would kids use it to support their collective learning or subvert already tenuous ties? Could the technology answer common questions and free teachers and kids up for higher-order tasks? Yes, I believe so. Could the technology, over time, help kids evaluate answers they get from AI, teachers and each other on their way to building critical thinking skills. Yes, I absolutely believe so. What questions are best for a peer…a classmate…a AI? At the same time, could the technology expose inequities and insecurities that kids at this age–really any age–have? Unfortunately, yes: “That’s such a stupid question!” “I literally just asked that.” “In my house…” As with hardware, teachers and administrators would need to establish clear expectations and consequences. If they did, and if they modeled productive behaviours, and if they included students in this process as part of ongoing lessons on digital citizenship, I believe that a product like Google Home, Amazon Alexa or Apple HomePod could add value and open up lines of communication that currently don’t exist. The flip side to this–one I can see critics teeing up–is that this will be yet another area where children interface with the virtual world and lose touch with the real world. It’s a consideration, to be sure, but I think that the unique combination of medium (voice vs. image) and the interface (shared vs. personal) are sufficiently compelling to explore further.


I expect that the 2017-18 school year could be the first one in which classroom-level AI could play a role. What role, though, and with what expectations? Google and Chrome have taken a leadership position in the education market, both hardware and software. Will this trend continue with AI? Can Google adapt or create an advertising model that makes class or group-level AI viable? (Related to this, are there other areas where group-accessible AI would make sense? Welcome centers? Hospital waiting rooms? Airports? DMS?)

Will Apple’s redoubled efforts in education–to wit they are making a return appearance to ISTE this year after a 10-year absence–mean anything in this area? Will Amazon just steamroll, delivering healthy lunches from Whole Foods and whole-child solutions? In the short term, would an AI speaker create more learning opportunities in class our would it be a burden to babysit?

“Hey, Siri…what role will smart speakers play in schools next year?”