Google has launched its AI chatbot, Bard, in more than 180 countries, including Australia, as part of the company’s annual I/O conference. Bard is built on Google’s large language model, PaLM2, and can provide information, write code, translate languages, and analyze images. Google plans to add visual responses to Bard in the future, and users will be able to upload images to be analysed by Bard using Google’s Lens application. Bard will also be integrated into Gmail, Docs, Drive, Maps, and other products as part of a feature called Duet AI, which will offer writing assistance within Gmail, Docs, Drive, and Maps.
Google has pledged to roll out the technology ethically and has included annotations on information sourced elsewhere to tackle the issue of AI hallucinations. Bard is currently available in English, Japanese, and Korean, and the company plans to make it available in more than 40 languages in the future. However, the company has said that systems built on PaLM2 “continue to produce toxic language harms,” which has delayed the launch in other languages.
As part of its promise to develop AI ethically, Google’s CEO, Sundar Pichai, said all of Google’s AI models would include the use of watermarking and metadata to allow people to know that AI-generated content is exactly that. The company would also add a new “about this image” tool in search results that provided context on where similar images might have first appeared and where else it was online. In addition, Google is limiting the use of the company’s new universal translator, an experimental artificial intelligence video swapping service that matches lip movements and translation with voice, to authorized partners due to the possibility of misuse by malicious parties.
Google’s AI chatbot, Bard, has been launched in over 180 countries, including Australia, and is built on Google’s large language model, PaLM2. The company has pledged to roll out the technology ethically, and as part of its promise to develop AI ethically, Google plans to include annotations on information sourced elsewhere and add a new “about this image” tool in search results. As part of Google’s new universal-translator AI video subbing experiment, it is limiting its use to authorized partners only.