Listen to the blog post here.

In the Sunday Blog post “A Brief History of Audio”, we went over some of the major audio inventions in history that overhauled the way we communicate. They were the telephone, the phonograph, and the radio. We then discussed how social audio continues the evolution of audio communication, but recent developments in technology have shifted people’s attention. 

Artificial intelligence (AI) is popping up in various aspects of our lives. So much so that people wonder how AI will affect everything, from jobs to healthcare to education. Inevitably, if not already, AI will also impact how we communicate. I want to focus specifically on how AI will impact audio communication.

• Speech recognition: AI-powered speech recognition technology is becoming increasingly accurate and can be used to transcribe and caption audio content. This can make audio content more accessible to people with hearing impairments. It also remedies a drawback of audio in that audio is linear and can’t be skimmed. Transcription makes it easier to search and analyze the content.

• Natural language processing (NLP): AI-powered NLP can be used to analyze the content of audio recordings, enabling machines to understand and respond to spoken language. This can be used in applications such as chatbots, virtual assistants, and voice-controlled devices. It can also help to summarize the key points of audio recordings, from short Voiceboxes to long podcasts and videos. 

• Audio synthesis: AI-powered audio synthesis can be used to create realistic-sounding voices for use in text-to-speech applications. This can then be used to create automated voice responses, personalized voice assistants, and even entire audiobooks.

• Noise reduction: AI-powered noise reduction algorithms can be used to filter out background noise from audio recordings, improving the quality of audio communication in noisy environments.

• Real-time translation: AI-based real-time translation technology can translate spoken words into different languages in real-time. The use cases for this abound, from international business to tourism to communication between people who speak different languages.

• Content moderation: One of the biggest challenges of social audio spaces is content moderation. Machine-learning algorithms can scan content at far greater speeds than human moderators can. This allows platforms to catch hate speech, adult content, or other types of content that should be either taken down or limited in reach. 

No doubt AI has a lot of promise for audio communication. However, like most technologies, AI does present some challenges to audio communication. For example, you may have seen people creating audio of one artist singing the song of another artist. While this may appear relatively harmless, the fact that AI can so closely mimic a person’s voice and get them to say anything has some dangerous use cases. 

Voice is often an identifying factor. From speaking on the phone with a family member to using voice-based security systems, people, pets, and other technologies recognize you by your voice. If AI can copy your voice, identify theft and its can of worms come into play.

We as a society will need to solve these challenges sooner rather than later. The rapid advancement of AI in the past year alone suggests that it will only continue to evolve and fine-tune. Whether in the voice space or otherwise, we must figure out how to leverage AI for good and overcome the challenges.

Tuesday Deep Dive is a series where we discuss in more detail a specific point made in the previous Sunday Blog.