With the SoundWave quickly spreading across the world and many brands already knee-deep in the voice-first era, there are many terms and definitions that have either evolved or are being used interchangeably. Voice only, voice-first, voice user interface, voice assistant, and voice AI are all used in specific contexts and refer to the technology, the strategy, or the end product and user experience. 

Two terms that have very different origins but are now often used almost  interchangeably are voice AI and conversational AI. While conversational AI began as a way to refer to the pop-up bots on customer service websites, its meaning has evolved to describe the most advanced and natural language voice user interfaces available.

As the voice AI industry continues to evolve at a rapid pace and more types of companies are joining the voice-first era everyday, we thought it would be helpful to help everyone understand the origin of some terms and their most common use cases with this helpful voice AI glossary of terms.

Voice AI

When talking about voice AI, we’re discussing the technology, the platform that makes the incredible capabilities of voice assistants possible. How do you use it in a sentence? Example: “SoundHound’s voice AI technology has high accuracy,” or “voice AI is growing in QSRs and retail.” It’s a broad term referring to the technology that powers devices—whether that’s in a TV, a kiosk, a drive-thru, a phone ordering system, or an in-car infotainment system. 

There are many components of voice AI, including 

• Automatic Speech Recognition (ASR): Software that is able to take audio input and map that input to a word or a language utterance.
• Natural Language Understanding (NLU):  A combination of components including a comprehensive lexicon, grammar rules, and semantic theory work together to break down sentences and guide comprehension.
• Text-to-Speech (TTS): Technology that converts text to audio that is spoken by the system.
• Wake Words or Wake Phrases: The spoken word or phrase that “wakes up” an always listening device.
• Content Domains: Data sets for content-rich conversational experiences
• Edge and Cloud Connectivity Options: From embedded-only to cloud-only and hybrid solutions.
• And more—all working together to create positive experiences for users. 

How is voice AI different from conversational AI? This becomes a little more complicated since the goal of voice AI is to be conversational. Verloop even referred to voice AI as a “conversational AI tool.” 

The differentiation may be in the technology itself. Voice assistants built on proven voice AI technology platforms deliver experiences that closely mimic the conversations one person might have with another. These advanced voice AI systems can understand complex and compound queries and are aware of the context of the conversation—allowing users to talk naturally without having to speak in short sentences, memorize scripts, or repeat themselves.

Conversational AI

The essence behind conversational AI is to be “human-like” or as close to interacting with a human as possible. Similar to voice AI, we’re referring to the technology, not the device itself. IBM states that conversational AI “refers to technologies, like chatbots or virtual agents, which users can talk to,” while TechTarget says, “Conversational AI is a type of artificial intelligence that enables consumers to interact with computer applications the way they would with other humans.”

While voice AI can be used across devices and industries, when we talk about conversational AI, we’re specifically discussing technology with enough processing power and advanced technology to allow users to be “conversational” when making requests. Conversational AI shows up in devices and services that either have access to the cloud, or a large enough CPU footprint to support the Natural Language Understanding (NLU) technology required to process the nuances of human speech and return fast, accurate answers to unscripted queries.

Some examples of conversational AI scenarios include:

Scenario: In-Car

• Query: Roll up the windows and navigate to the nearest gas station.
• Answer: All the windows are up. There is a Best Gas in one mile. Do you want to go to that one?

Scenario: Airplane passenger

• Query: Call stewardess for a glass of soda and send a message to John Smith my flight is delayed, I will be at Glasgow airport at 8pm
• Answer: A crew member will bring you a glass of soda. Message sent to John Smith from your contacts: my flight is delayed. I will be at Glasgow airport at 8pm.

Future advancements in the field, like sentiment analysis, could broaden the horizon and bring conversational AI even closer to that “human-like” state. 

Conversational intelligence

A slight variation of conversational AI, conversational intelligence has traditionally been used when describing chatbots for a customer service purpose. Authenticx has described conversational intelligence as “software that uses artificial intelligence (AI) to analyze speech or text in order to derive data-driven insights from conversations between sales agents and customers.”

While this was true when chatbots were the primary AI interface between sales agents, contact centers, and help centers, it’s not longer the case. The term conversational intelligence is evolving along with voice AI technology to signify the machine learning component of voice AI. 

The same technology that began using AI, machine learning, and natural language processing technology to draw meaning from unstructured data to answer questions and enhance customer support is now responding to voice input and responding as a human would in applications both inside and outside of the call center or customer support application. 

Voice assistant

Now that we’ve parsed the meanings between voice AI and conversational AI, we can move on to the variations of voice, beginning with voice assistant. A voice assistant is the actual interface between the device and the user.

Originally associated with the in-home speakers developed by the big tech players, the term voice assistant is now used more generically to describe the anthropomorphized “person” that greets you when you wake up a voice-enabled device.

According to Slang Labs, “Voice Assistant is a virtual assistant that uses speech recognition, natural language processing and speech synthesis to take actions to help its users. These assistants have evolved quickly and can perform several complex tasks today.”

While big tech is primarily focused on voice assistants in smart speakers, companies across industries are implementing wholly-owned voice assistants to voice-enable a wide range of devices, services, and apps. 

A voice assistant can also be referred to as a voice interface or voice user interface. These terms are considered interchangeable and can be used to describe the same entity.

Voice experience

A voice experience—whether good or bad—is the result of the interaction between the user and the voice assistant. For example, the voice experience is often improved by the availability of certain voice AI technologies, such as context-aware or voice ID. For brands and developers, the goal of implementing a voice assistant or partnering with a voice AI platform provider may be to design an exceptional voice experience that will lead to greater customer loyalty and brand satisfaction.

In summary, as voice AI technology evolves, terms like conversational AI, conversational intelligence, and voice AI will continue to morph into terms that describe a variety of voice experiences. Regardless of how any one voice AI solution is named or described, the keys to a voice AI solution that will solve today’s challenges and lead your company into the future remain the same:

• Branded voice experiences
• Ownership of the customer experience
• Data insights and transparency
• Customizable solutions

Want more breakdowns of voice AI technology? Check out the following: 

• What You Need to Know About Wake Word Detection
• How SoundHound’s Voice AI Technology Delivers Conversational AI That’s More Human™
• How to Choose the Conversational AI Strategy Best for Your Brand and Users
• Discover the Possibilities of Voice Assistants on the Edge

At SoundHound, we have all the tools and expertise needed to create custom voice assistants and a consistent brand voice. Explore SoundHound’s independent voice AI platform at SoundHound.com or speak with an expert or request a demo below.