South Korean users of voice-controlled smart speakers are largely dissatisfied with the products over issues such as failing to recognize voice commands, a survey showed Friday.

Voice-controlled smart speakers are artificial intelligence-based devices that respond to voice requests from users and provide relevant services. The AI-speakers that are currently in the domestic market are KT’s Giga Genie, SKT’s Nugu, Amazon’s Echo and Google Home. 

The Korea Consumer Agency’s survey asked 300 users of the smart speakers about their experiences using them. They were allowed to pick multiple choices.

According to the survey, 56.7 percent of the respondents said that they found the inbuilt voice recognition systems had difficulty understanding their voices.

Some 45.7 percent said they had trouble communicating with the AI systems, and 37 percent said the speech-recognition systems recognized noises as commands.

“Customers may have high expectations of the voice-controlled smart speakers since they are often advertised as being equipped with high-end AI technology,” the KCA said. It is important for the developers to notify customers about such technological issues, the KCA added.

Meanwhile, the survey also showed that the most frequently used speech-recognition-based function was playing music, at 71.3 percent. The second most popular function was weather and traffic information services with 41 percent. Internet searches and timer functions followed with 40.3 percent and 35.7 percent, respectively.

When asked why they purchased such devices, 67.7 percent answered “curiosity about AI-based devices.” 

By Shim Woo-hyun (ws@heraldcorp.com)