Google Cloud held an important launch conference at its DeepMind headquarters in London, officially launching its latest high-definition voice model Chirp3. This model is open to developers through the Vertex AI unified machine learning platform, providing a wealth of development tools designed to promote program innovation. The launch of Chirp3 marks another major breakthrough for Google in the field of voice technology.
Chirp3 supports 248 different sounds and can be synthesized in 31 languages. Developers can use this model to create a variety of applications, such as smart voice assistants, audiobooks, and video dubbing. Google says Chirp3's voice feature can capture nuances in human voice, making the conversation more vivid and engaging. The application prospects of this technology are broad and will bring users a more natural and smooth voice interaction experience.
In addition to using ready-made voice, users can also create custom voices through Google Cloud's text to voice API. However, to ensure responsible use, Google restricts access to this voice cloning feature to prevent potential abuse and ensure ethical AI practices. This move reflects Google's emphasis on ethics and responsibility in technology development.
At the press conference, Google Cloud CEO Thomas Kurian emphasized that Google's overall vision is to provide a wide range of models, including Gemini, Imagen, Veil and other products in addition to Chirp3. Google also launched a new product called Agent Space, designed for business users to meet their specific needs. The launch of these products will further enrich Google's AI ecosystem and provide users with more choices.
Sir Demis Hassabis, CEO of Google DeepMind, also introduced Gemini's evolution, especially its multimodal understanding. He mentioned that in AI Studio, users can post YouTube links, and Gemini can process video content, use its long-term context window to let users ask questions, and quickly find critical moments for lectures or sports events. This feature will greatly improve the efficiency and experience of users when processing video content.
In addition, Google has taken the opportunity to announce a plan aimed at improving the UK's AI skills through comprehensive training programs and helping professionals effectively master AI technology. Google will provide UK startups with points for the use of cloud infrastructure and AI tools, supporting them to quickly develop and scale innovative solutions to stimulate entrepreneurial activities. This plan will help promote the development of the UK's AI industry and cultivate more AI talents.
In terms of privacy and compliance, Google reiterates its commitment to data residency, and its Vertex AI and Agent Space AI tools can help organizations train and serve models in compliance with local laws. This is crucial for industries such as healthcare and finance, where privacy and compliance are extremely demanding. This commitment from Google will increase users' trust in their products.
Project: https://cloud.google.com/text-to-speech/docs/chirp3-hd
Key points:
Google Cloud launches the Chirp3 voice model, supporting 248 sounds and 31 languages, helping developers build smart applications.
Google restricts access to voice cloning capabilities to ensure ethical AI practices and prevent abuse.
Google initiates plans to enhance UK AI skills and provide cloud infrastructure support to start-ups to promote innovation.