Voice on Voice

While CES was a shadow of its typical self this year, the news coming out of Vegas still got technology enthusiasts like us excited about what the coming decade has in store. The conference showcased a diverse array of tech, from color-changing BMWs (complete with the surprising re-birth of e-ink technology), to aggressively curved 55” Samsung computer monitors (the new pandemic WFH luxury?), to an at-home hair-coloring alternative from L’Oreal to keep you from omicron at the salon.

There’s a saying, coined by the American author William Gibson: the future is already here, it’s just not evenly distributed. Today’s batch of news demonstrates how voice technology is being distributed across new spaces, whether it be the web, motorbikes, or glasses.

The ability for a computer to translate speech into text, and text into meaning, is still in its early days. But to improve, it has always relied on data. Whether that’s come in the form of academic contests and active solicitation of human feedback, or using machine learning to passively recognize sounds like running water, voice as interface is always getting better, the more we interact with it.

Historically, the voice assistants of Alexa, Google Assistant, and Siri have been female by default — spurring conversation on the gender biases and stereotypes that these voices perpetuate. Voice assistants were created to perform tasks for us, and by making them all female, it subtly reinforces antiquated social constructs. Tech companies are actively trying to combat this by introducing new voices or changing the default voice for their assistants.

While Apple, Google, and Amazon have spurred the rise of mainstream voice assistants, custom voice assistants now have more share of the conversation. Companies including Nuance Communications and SoundHound — whose platforms enable custom assistants — have enjoyed great success in the past few years, either being the target of a major acquisition or going public

One of the major pain points of voice assistants is their inability to understand us at times. For those with speech impairments, voice technology has been largely inaccessible. The industry is actively exploring ways to make assistants more inclusive. Last year, startup Voiceitt closed $10 million in Series A funding to build software that can understand those with atypical or impaired speech. Amazon partnered with Voiceitt to improve Alexa.

From Siri to Alexa to Google Assistant, people use a variety of voice assistants in their daily lives, depending on what device they’re talking to. However, this creates friction in the household — that’s why initiatives like Matter and Alexa Custom Assistant are exploring how smart home devices can house multiple assistants at once.

Interactive voice advertisements have caught the attention of major brands looking to tap into novel ad experiences. Attribution has long been a pain point for advertisers — but with interactive audio ads, brands know when a listener engages with their spots, providing valuable insight into the effectiveness of a campaign.

A few weeks ago, The Verge explored Siri’s advances since its launch 10 years ago. While it was the first major voice assistant to be introduced, it has fallen behind competitors Alexa and Google Assistant, lacking complex functionality beyond answering questions and sending texts.

The past year has seen a flurry of voice activity in the QSR space, particularly in the drive-thru. White Castle, McDonald’s, and this week, Wendy’s, are all deploying voice to take customer orders. While voice ordering isn’t a novel use case, these implementations illustrate how valuable voice can be when paired with complementary technology.