Google chatbot Bard gets AI fact checker

In the race between artificial intelligence systems, Google initially lost out to ChatGPT. But the internet giant is constantly improving its chatbot Bard, which is co-developed in Zurich. However, some innovations will initially only be available in English.

In the future, Google’s chatbot Bard will help users recognize false facts in the AI ​​system’s responses. Bard’s answers can be checked with a click on the button with the colorful Google logo, Google CEO Jack Krawczyk announced on Tuesday in Mountain View.

The parts of the answer that the Bard dialogue system is very confident about are then highlighted in green. The passages in which Bard also discovered information on the internet that could refute this statement are colored orange.

Krawczyk said Bard is the first AI language model that is willing to admit when it is unsure of something by not highlighting it in green. “Bard goes even further and says, ‘I am willing to admit that we made a mistake.’” The ‘double check’ will be made available worldwide, but initially only in English.

Bard switches to Gmail, Google Drive and Co

In another innovation, Google wants to enable users to connect their Google services such as Gmail, Google Drive, Flights, etc. with Bard. For example, you can use Bard to search for email or Google Drive content or plan a trip including flights and hotels.

In the German-language version of Bard it will also be possible to combine voice input with image uploading in the future. For example, you can upload a photo of the label of a wine bottle and have Bard explain in detail which main course suits it best. For example, it is also possible to “have Google Lens show the newly purchased child seat and have Bard search for a suitable YouTube video on how to install it in the car,” writes the German tech portal heise.de.

The feature to share a Bard chat history with other people is now also available in German.

Google emphasizes that emails and Google Drive documents read by Bard are not viewed by human reviewers, are not used for advertising purposes, and are not used to train the Bard model.

Google initially reacted very hesitantly to the initiative of the Californian start-up OpenAI, which had more than 100 million users in just a few weeks since November last year with its text robot ChatGPT. Initially, Bard launched in English only in the US and UK in mid-March. Bard is now available worldwide in 40 languages.

(oli/sda/awp/dpa)