Google Launches AI-Powered Text-To-Speech Advancements
Cloud Text-To-Speech sees Google applying AI and neural networks to make speech processing 1000x faster.
Google is rolling out a new AI-backed voice service, Cloud Text-To-Speech, aimed at enabling any developer or business to synthesize speech on demand — whether for an app, website, or intelligent assistant.
Backed by WaveNet (which is powered by DeepMind, the AI-focused company Google purchased in 2014), the move sees Google doubling down on applying advances in AI to creating more seamless voice-based interactions — something that makes sense as consumers’ voice searches continue to skyrocket.
Per Google’s statement, “Cloud Text-to-Speech enables developers to synthesize natural-sounding speech with 30 voices, available in multiple languages and variants” — meaning that it’s now easier to create authentic interactions with users across devices.
The ‘Voice Future’ Is Now
While this update is a bit complex, spotlighting Google’s investments in both AI and building neural networks, it’s easy to see why it matters to marketers at its core: As more consumers interact with their devices via voice, it behooves brands to develop a voice strategy now, rather than lose out on interactions — and, potentially, on becoming a voice assistant’s top answer to searchers’ queries.
As Sherry Bonelli wrote earlier this month, “according to ComScore, by 2020, 50% of all searches will be voice searches. People are changing the way they search. We’re getting lazy and touching those tiny keypads on our smartphones is just too much work – so more and more of us are using our voice and normal every-day language to ask search queries to our computers, smartphones, tablets and smart voice assistants.”
And with WaveNet’s model backing Cloud Text-To-Speech now reportedly 1,000 times faster at processing speech — producing 20 seconds of higher quality audio in just one second — the future of voice has already taken a big leap forward.