From Music to Mind Reading: AI Startups Bet on Earbuds

Startups have been trying to expand on what headphones do, well before a large number of startup companies began to release them, such as Waverly Labs and Mymanu, which launched their real-time translation headsets almost a decade ago. This was an indication that the earbud was entering a new phase of development, towards more intelligent and capable earphones. Google followed suit with the launch of its voice AI assistant in 2020, while other major tech companies such as Samsung and Apple introduced noise-cancelling technology, which has become a “trademark” of high-end audio products.

Today’s new generation of startups is proposing and developing New uses of artificial intelligence technology and will feature many of these advances at the annual CES show in Las Vegas.

OSO is one such startup, who is working to turn their earbuds into something more useful as a personal assistant. Its devices can record meetings and retrieve key parts of conversations on demand, using natural, everyday language.

A competitor, Viaim, offers similar functionality but is focusing on interoperability in an ecosystem dominated by major smartphone makers and their proprietary platforms.

“If you use a different brand of phone, you often lose access to AI features altogether,” said Viaim CEO Shawn Ma. “That’s where our earbuds come in.” Viaim’s products are designed to work across brands, including with iPhones in China.

Meanwhile, Timekettle has found success in a very different niche. According to Brian Shircliffe, head of U.S. sales for the Chinese company, about 90 percent of its sales come from schools. Many educational institutions provide the earbuds to non-English-speaking students, allowing them to follow lessons without needing a human translator.

Reading minds
Whether earbuds could eventually replace smart glasses, connected speakers, or even smartphones as the primary interface for generative AI remains an open question. For now, most AI-powered features still depend heavily on the smartphone they’re connected to, said Ben Wood, chief analyst at CCS Insight.

“Earbuds are a more accessible entry point for AI than smart glasses,” said Avi Greengart, president of consultancy Techsponential. They’re cheaper, widely adopted, and don’t require prescriptions. However, he noted that people don’t wear earbuds all the time, unlike glasses, and voice-based interaction only works in environments where speaking out loud is appropriate. The lack of a camera also limits their potential.

Some startups are working around those constraints. An example of the possibilities created by advanced technology can be found in Naqi Logix, which has developed Neural Earbud devices. The Neural Earbuds are equipped with advanced ultra-sensitive sensors that detect very slight movements in non-visible areas of the user's body; thus, allowing a quadriplegic person to use the device to navigate either a wheelchair or a computer by merely gazing at a particular screen.

Operations manager Sandeep Arya is optimistic about the future of these innovations: "There is a strong desire among people to interact more subtly and quietly with their environment. They do not need to make vocal commands to Siri, Alexa, and other AIs," he said.

Arya appears to believe that as the sensor technology continues to improve, additional future versions and applications will include interpreting facial contractions and emotional states so that the tone, pitch and style of a Chatbot can be adapted accordingly.

Another new startup taking a different approach to the use of Digital Humans is Neurable, whose mission is to develop and apply technologies for understanding emotions.Its MW75 Neuro LT headset measures brain activity, with the long-term goal of enabling communication through thought alone, without words or gestures.

“It’s remarkable,” said Ben Wood of these developments, “but for now, it remains a niche market.”

Until then, the hundreds of millions of headphones already in circulation will continue to focus primarily on what they do best: listening.