Google’s AI chatbot Bard launched for Australian users on Thursday as the company showcased its advancements in artificial intelligence and pledged to roll out the technology ethically.
Until now, Bard was only available in the US and the UK, but on Thursday at the company’s annual I/O conference Google announced it would open up the chatbot to users in more than 180 countries around the world, including Australia.
Bard is the chat program built on Google’s large language model, PaLM2, similar to how ChatGPT is built on OpenAI’s GPT. It can provide information, write code, translate languages and analyse images.
As part of future advancements to Bard announced by Google on Thursday, Bard will provide visual responses in addition to text-based responses. Using Google’s Lens application, in the future users will be able to upload images to be analysed by Bard. Google used the example of the photo of two dogs with the prompt “write a funny caption for these two” and Bard will be able to determine the breed of dogs and draft responses.
In a bid to tackle the issue of AI hallucinations – whereby the AI creates a sourced text or information that it claims to be true – Bard will include an annotation on the information sourced elsewhere and provide a link to it.
The company is also working to make the chatbot available in more than 40 languages. Currently it is available in English, Japanese and Korean. The reason for the slow launch in other languages is that Google has said that based on preliminary research, systems built on PaLM2 “continue to produce toxic language harms”.
Google also plans to integrate Bard into Gmail, Docs, Drive, Maps and its other products over time in a feature it calls Duet AI that will allow users to get assistance with writing and other work functions within the apps. But the company has stressed that users will be in control of their privacy and how the tools are integrated into these products.
As part of Google’s promise to develop AI ethically, the company’s chief executive, 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.
Included in future advancements in search that will see AI-generated text results along with links, the company would also allow 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.
Google is also allowing only authorised partners to use its new universal-translator, experimental AI video subbing service that translate’s a speaker’s voice and matches their lip movements. The company said while it had “enormous potential”, there was a large risk of misuse in the hands of bad actors.
