The realm of generative AI has largely revolved around text and image creation, but the horizon is expanding as voice technology gains momentum. A recent exciting development sees Google introducing Chirp 3, an advanced speech-to-text and high-definition text-to-speech model, integrated into its Vertex AI platform starting next week.
Building on this progress, Google announced that Chirp 3 would now offer eight new voices available in 31 languages. This expansion opens up numerous possibilities, such as developing voice assistants, creating immersive audiobooks, building support agents, and generating voice-overs for a range of multimedia. This thrilling update was unveiled at a Google event held at the DeepMind offices in London.
Meanwhile, other key players in the AI voice space are making significant strides. For instance, Sesame, the creators of the highly realistic sounding AI personas “Maya” and “Miles”, have launched a model enabling developers to craft personalized apps and services using their technology.
Despite these advancements, Google is mindful of potential misuse. Thomas Kurian, CEO of Google Cloud, mentioned the need for specific usage restrictions while collaborating with their safety team, reflecting a careful approach to technology rollout.
Among companies delving into AI voice services, startups like ElevenLabs have attracted substantial funding to fuel their innovations. Chirp 3 joins other cutting-edge AI models such as the next iteration of Google’s flagship language model, Gemini, the image-generation tool Imagen, and the sophisticated Veo 2 video creation software.
Whether Chirp 3 reaches the level of realism seen in projects like Sesame remains to be seen, but as DeepMind CEO Demis Hassabis pointed out, the journey with AI is a marathon and not a sprint. Hassabis shared his perspective, projecting AI’s transformative impact over the next decade rather than in the immediate future.
The unveiling of Vertex AI in 2021 was a strategic move by Google to empower developers in building cloud-based machine learning services long before the AI frenzy ignited by OpenAI’s GPT initiatives. In response to competition from tech giants like Microsoft and Amazon, Google has focused on enhancing Vertex AI, offering capabilities that enable developers to classify data, train models, and transition these models into production environments.
Furthermore, Google has nurtured its “Chirp” voice services for several years, initially as a contender to Amazon’s Alexa. With these advancements, it will be intriguing to see how Google continues to expand its technological ecosystem while possibly integrating models beyond its proprietary creations.
The ongoing evolution in AI voice technology promises an exciting future, potentially revolutionizing how we interact with machines and content in our daily lives.
