Microsoft has hit a milestone in voice recognition

Voice Activated Personal Assistants
Voice Activated Personal Assistants

(BII)
This story was delivered to BI Intelligence Apps and Platforms Briefing subscribers. To learn more and subscribe, please click here.

Microsoft Research developed a voice recognition system as accurate as humans, with an error rate of just 5.9%, according to Microsoft's Chief Speech Scientist Xuedong Huang.

The historic breakthrough comes just one month after the company surpassed IBM Watson's previous record of 6.9%. Microsoft will likely incorporate the improved system into its various products including voice assistant Cortana, Xbox, and other productivity tools. These advancements will also spur competition among tech companies, which will likely boost overall voice recognition capabilities.

Rapid advancements in voice recognition accuracy is going to help the technology emerge as a dominant computing interface. That's because it's much faster and easier to talk to your device than type on it. A recent Stanford University study showed that voice input is three times faster than keyboard input, and with a 20.4% lower error rate.

And in an attempt to move beyond the hype, tech companies including Amazon, Apple, Google, and IBM, are already deploying voice assistance to consumers. Apple, for instance, recently added third-party integration to Siri, which will enable it to communicate with other apps. This means that users can order an Uber through Siri. And Google is slowly rolling out Google Assistant to its users.

But voice assistants need to overcome a number of hurdles before mass adoption occurs:

  • "As close as humanly possible" is not good enough for voice assistants. Despite the impressive results Microsoft's research yielded, speech recognition needs to reach roughly 99% for voice to become the most efficient form of computing input, according to Kleiner Perkins analyst Mary Meeker. This is likely because expectations for automated services are much less forgiving than human error allows. In fact, when asked what voice assistants could do better, "understand the words I am saying" received 44% of votes, according to MindMeld.

  • Consumer behavior needs to change. While the use of voice assistants is increasing, many users are still uncomfortable speaking to their assistant. Thirty-four percent of millennial respondents in an exclusive BI Intelligence survey were either unaware of voice assistants, or unwilling to use them freely. And 18% said they would use Amazon's Echo only if they weren't around other people. Millennials surveyed by BI Intelligence skew toward the tech-savvy, which means that the level of discomfort among millennials in the general population is likely higher. For voice to truly replace text or touch as the primary interface, consumers need to be more willing to use the technology in all situations.

  • Voice assistants need to be more helpful. Opening up third-party apps to voice assistants will be key in providing consumers with a use case more in line with future expectations of a truly helpful assistant. Voice assistants like Siri, Google Assistant, and Echo, are only just beginning to gain access to these apps, allowing users to carry out more actions like ordering a car.

To receive stories like this one directly to your inbox every morning, sign up for the Apps and Platforms Briefing newsletter.Click here to learn more about how you can gain risk-free access today.



More From Business Insider

Advertisement