[Note: This post by Amy Stapleton recently appeared on LinkedIn and is of such relevance and import that we are sharing it here.]

What’s Stifling Voice Adoption, or Why the “Voice App Gap?”

Recently there’s been a discussion in the voice community that boils down to the question: why are we so underwhelmed by voice experiences? Kane Simms led an interesting discussion on the topic on VUX Voice World with Dustin Coates, John Kelvie, Ahmed Bouzid, and Bret Kinsella.

Bret Kinsella has addressed the topic of what he terms “the voice app gap” in several of his latest Voice Insider newsletters (highly recommended!).

Here are my thoughts.

Our Designs Are Close to Catching Up to Technology

Perhaps our biggest stumbling block really was technology related; or maybe I should say the challenge has been in implementing voice apps/products that take full advantage of the very latest tech (especially the ability to quickly find, scan, and summarize content and then quickly generate dialogue based on the “understanding” gleaned from those content sources.).

I also think the real breakthrough is tied to a rethinking of the “intent based” model of conversational AI that I criticized from the beginning.

Use Case for the Next-Generation Voice App – My Morning News Routine

I enjoy catching up on the latest tech news over breakfast. I read news on Business Insider and a couple of other tech news sites. I scan the headlines, skipping most articles, and picking 2 or 3 to dive into. There is currently nothing available in the voice world that would make me switch up my routine. I don’t want to listen to the Business Insider flash briefing. It gives me too much info that I don’t care about. It’s not an enjoyable experience. I don’t want a TTS engine to read me articles. I’m much better at just browsing, reading, or skimming.

A Voice Experience That Would Make Me Give Up My Morning “Read”

If the following simulation really worked, I would use it. Imagine me sitting down to my breakfast. Either in an audio-only experience or by engaging with a virtual character, I start the following conversation.

Simulation

Me: What did I miss?

Bot: Did you hear Jeff Bezos stepped down as CEO of Amazon?

Me: What? Are you serious?

Bot: Yep, it’s true. He made the surprise announcement when the company released its latest financial reports. He says he’s stepping down in three months, but he’ll remain the executive chair of Amazon. Andy Jassy will take over the role of CEO. Do you know who Andy Jassy is?

Me: I don’t think so.

Bot: He’s currently in charge of the AWS division. He joined Amazon a long time ago and he’s been one of Jeff Bezos’s top confidants. [Pause]

Me: Wow, I’m really surprised. [Pause]

Bot: Yeah, it seems lots of people were surprised. Bezos says he plans to spend more time on his other endeavors. He listed Blue Origin and The Washington Post. [Pause]

Me: Ok, so…. does that mean he’ll still be in charge at Amazon?

Bot: Technically he won’t be in charge, but as executive chair he’ll have a lot of input into key decisions. [Pause]

Me: So do people really believe his reasons for stepping down or is there more to it?

Bot: There’s speculation that the decision may have something to do with the impending antitrust investigations that are likely to happen in the new administration. Bezos was called before the House Judiciary Committee last year, where he faced questions about the company’s competitive practices, including accusations of pricing out competitors on its platform and using third-party seller data to inform the sale of its own products. If Bezos isn’t CEO, it’ll probably take the spotlight off of him. [Pause]

Me: Is that credible information or fake news?

Bot: Well I got the information from (cites one or more sources). [Pause]

Me: Ok. I’ll take it with a grain of salt. [Pause]

Bot: Understood. [Pause]

Me: So what are the other reasons people say he might be stepping down?

Bot: Do you want me to stick to the mainstream, or do you want to hear the wild conspiracy theories?

Me: [Laugh] Oh, what the heck, let’s hear the wildest theory you can find. [Pause]

Bot: The wildest one I’ve found so far is that aliens are holding the real Bezos hostage and refused to release him until he stepped down as CEO. [Pause]

Me: [Laugh] Where did you get that from?!

Bot: Cites source. [Pause}

Me: Got it. When did you say he’s stepping down?

Bot: In three months. [Pause]

Me: Hmm. I’m still really surprised. How much of Amazon does Bezos own?

Bot: It says he owns over 10%, or something like 53 million shares. [Pause]

Me: Sheesh. I guess he can do whatever the heck he wants. [Pause]

Bot: Yup.

What Technologies Are Missing To Make This Experience Work?

None. We have all the pieces required to build a conversational newsbot that works as described above. Based on Kinsella’s interview with Igor Jablokov, his company Pryon is already delivering a similar experience for enterprise customers.

Granted, building a powerful conversational bot like the simulated newsbot is not easy; but I believe the value of these smart conversational AI assistants is so self-evident that we’ll see the market shift in this direction.

(I know Quartz was doing something vaguely similar to what I describe with their news chatbot, but it was nowhere near what I’m envisioning in terms of its ability to converse in a natural and personalized way).

Just Think of the Value!

Consider how much value you could get from having access to such an AI assistant. It’s not just about saving time catching up on the latest news. The AI can evaluate far more sources of information than you could reasonably read. If instructed to do so, the bot could bring you different viewpoints on the same subject and potentially help you decide which has the most merit.

The use cases go well beyond helping you with breaking news. The same conversational AI framework and tech could help you get smarter on everything from current science, health and wellness topics, traffic and weather. It could help us evaluate and select products and services, manage our finances, and excel in our careers.

The Sun Has Not Set on Voice

Technology has given us the ability to: understand and generate natural language; ingest structured and unstructured information sources and accurately summarize the knowledge they contain; and answer specific questions about the information.

I’m optimistic that we’re getting very close to a world of smart assistants that can find information for us, read through it and “understand” it, tell us about it and answer our questions, and tailor the information they bring to us based on our preferences–all in a very conversational manner.

Let’s keep building!

Categories: Intelligent Assistants, Articles