Maury understood her frustration. He, too, was losing his vision and could foresee a struggle to remain productive using the visual interfaces common on all computing devices.
Conversational computingThe woman’s wish serves as a kind of voice command for Maury, whose company, Conversant Labs, is trying to lead other developers in the direction of computing by voice. His company’s latest creation, an app called Yes Chef, is a recipe app for iOS that can talk blind people through a recipe step-by-step.
While Yes Chef was made with the blind in mind, it’s for anyone who cooks. Rather than stopping to wash dirty hands to scroll down a device screen, a person can just ask the app for the next step — or even repeat previous steps — in the recipe.
Confused about whether to use baking soda or baking powder? Yes Chef will answer the question. It won’t chop your vegetables, but it could tell you how finely to chop them if it’s important to the recipe.
“Our mission is to make voice a meaningful alternative to visual computing,” said Maury, founder of the Pittsburgh-based company. “When we got started, we wanted to solve problems for the blind, but we wanted to make things useful for the general public. The best scenarios for accessibility is an experience that is the same for everybody and it’s just accessible by default.”
Yes Chef hits the App Store (it debuted on Aug. 14) at a time when developers and computer companies are creating more software with conversational user interfaces and pushing voice-enabled computing as a mainstream platform.
There are several big players in the field of voice-enabled technology. In addition to Apple’s Siri, there is Amazon’s Alexa paired with Echo smart speakers. Microsoft has Cortana and Google recently announced Google Home.
Amazon invites developers to write software to give Alexa and the Echo greater range. Apple is also said to be working on a device to rival the Echo and recently announced it would release an SDK (software developers kit) for creating apps that work with Siri.
Knowing what you mean
Yes Chef showcases Conversant Labs’ work on advanced algorithms that offer a more intelligent understanding of the meaning of words.
Maury gives this example: You can tell Siri you are hungry for nachos and it will understand and might respond with a website that lists recipes or offer the names of restaurants in your area. If you say to Siri, “I have the munchies for nachos,” it will understand the words but not recognize the sentence as a request for help in finding nachos.
Yes Chef handles more-complex interactions and even understands subtle and casual language with only a slight learning curve, Maury said.
“Two or three years ago, this might have seemed gimmicky,” Maury said. “On the technology side of things, there were no tools or libraries for building voice-enabled applications. We have a very firm grasp on what it takes to build a responsive conversational application. We’re trying to make that expertise available.”
Developers can download an SDK from Conversant Labs.
Using the voice for everything
Maury was a product manager for tech startups in 2011, when he was diagnosed with Stargardt macular degeneration and told he would be blind within 10 years.
As he prepared, he discovered there were limited tools for him to remain productive once his sight was gone. In 2014, he started Conversant Labs. Its first product was a shopping app that gave the visually impaired the ability to order items through voice commands.
Maury volunteered at a senior center in Pittsburgh to teach visually impaired people how to use their devices and gauge how to simplify the computing experience through voice.
“You want to be able to tell the computer to do something, understand you, and fulfill your wish,” Maury said. “The Echo and Siri are great but they have limitations. With Siri, you can identify a song or order an Uber, but it can’t read you the news. We want to create something that supports everything.”
Maury is confident consumer-driven devices will be available and understand complex requests within the next couple of years.
The next frontier for voice computing will be the workplace and finding ways to easily manage email by voice.