The recent news of Amazon reducing staff and budget for the Alexa department has put the voice assistant industry under scrutiny with many believing it is stagnating or even in decline. However, this idea is flawed because though voice technology has hit its use-case ceiling, it does not mean it is stagnating. It only implies that the current technology has several limitations that need to be addressed to support further evolution.
Superior natural language understanding (NLU), voice metadata extraction, and overcoming crosstalk and untethered noise are the three capabilities required to meet the human standard. Though many companies have conquered the first two aspects, the third capability remains the ceiling for current voice technology.
Achieving these three capabilities will give a voice technology functionality that could transform industries such as the restaurant and hospitality sectors. Currently, the use of voice assistants is limited to answering simple queries that do not require meaningful interactions. However, specialized, futuristic commercial deployments require voice assistants or interfaces to have use-case specialized capabilities like human-like interaction with the user, voice metadata extraction, and cross-talk resistance in public places.
To further break through the ceiling, technology needs to overcome untethered noise challenges and crosstalk. It not only needs to capture the voice of interest but also extract metadata in voice such as biomarkers. Extracting metadata could open up voice technology’s ability to understand emotion, intent, and mood. In addition, voice metadata could permit personalization.
To enable useful voice control, voice technology must be tested in real-world scenarios and deployed in specific environments with standardized benchmarks it must meet. With these actions, wider adoption of voice technology across industries is possible.
In conclusion, the recent announcement from Amazon is not a signal of the voice technology industry’s decline. Instead, companies need to address the current limitations of voice technology and continue to push for innovation that meets the human standard.