We are headed toward a convergence of audio technology and artificial intelligence (AI). As podcasts, audio books, and voice-operated devices gain popularity, artificial intelligence is becoming faster and more accurate.
Right now, we don’t really have technology for real-time audio interpretation. Live audio broadcasts cannot be censored in real-time because humans are not fast enough to interpret audio and censor it before it reaches listeners. Current live television broadcasts are actually delayed by seven seconds in order to give censors enough time to censor unacceptable content before it reaches viewers.
Where humans might need seven seconds, artificial intelligence computers might need only a fraction of a second, or even anticipate what is about to be said. AI may finally make real-time captions, censorship, and transcription possible. In fact, some computer scientists are already working on it. For example, Intel now has software for real-time hate speech censorship. A number of companies have released software that automatically captions audio. Google has AI-powered speech-to-text software.
Real-time audio interpretation is still in its infancy, because the artificial intelligence that makes it possible is also in its infancy. We will definitely see a lot more AI-powered audio tools in the near future. If you are a software developer, this is an area that shows a lot of promise for the next “killer app.”
Leave a Reply