Google’s conversational AI application, Bard, is finally about to enter public beta. Today, Google announced that Bard is accepting user applications from the UK and US markets, and over time, Bard will support more markets and languages.
Currently, users who sign up for Bard are placed on a waitlist, and Google will gradually open up testing permissions to registered users. However, it seems that no users have been granted permission yet.
Google claims that users can use Bard to improve work efficiency, speed up idea conversion, and stimulate curiosity, such as explaining quantum mechanics in simple terms or asking Bard to help create a reading plan.
In the initial testing for invited users, Google has already improved Bard, and will continue to launch it to more users and obtain more feedback to improve it.
Bard is a conversational AI application supported by LLM, a large language model. The application also uses a lightweight and optimized version of the LaMDA model, and Google will continue to update and provide more powerful models over time.
Based on Google’s understanding of high-quality information, Bard can be seen as a prediction engine that predicts the follow-up content of user input. Google has adjusted the flexibility of Bard, so the content input by users may generate more creative ideas by Bard.
Google emphasizes that although LLM is an exciting new technology, it is not without its flaws. For example, because they have learned various content, including some negative content, this may lead to LLM presenting inaccurate, misleading, or even false information.
Despite these issues, LLM still has incredible benefits, such as quickly starting human productivity, creativity, and curiosity. Therefore, when using Bard, users can choose several different response drafts and select the content they need based on their actual needs.
Application address: http://bard.google.com/