To surpass ChatGPT in popularity, Alphabet Inc.’s (GOOGL.O) Google announced on Tuesday that Bard, its generative artificial intelligence, will be able to fact-check its responses and examine users’ individual Google data.
Last year, the launch of ChatGPT, a chatbot from OpenAI supported by Microsoft (MSFT.O), spurred a scramble in the tech sector to provide users with access to generative AI technologies. ChatGPT, which is currently among the top 30 websites in the world, was the consumer application with the quickest growth rate ever at the time.
Bard has not experienced the same success. According to online analytics company Similarweb, it garnered 183 million visits in August, 13% more than ChatGPT.
Google is releasing Bard Extensions, which lets users input their data from other Google products, as it looks to gain a position in the quickly evolving AI market. Users might ask Bard to look up specific files in Google Drive or to summarize their Gmail inbox, for example.
Users of Bard may currently only access data from Google apps. Still, Google is collaborating with outside businesses to integrate their applications into Bard in the future, according to Google Senior Product Director Jack Krawczyk.
Another new Bard feature aims to solve a persistent issue with generative AI: false positives, sometimes known as “hallucinations.” Users of Bard will be able to determine which details of Bard’s responses conflict with and concur with Google search results.
By holding Bard accountable, Krawczyk added, “We are presenting (Bard) in a way that it admits when it’s not confident” to increase users’ confidence in generative AI. Thanks to a third new feature, users can invite others to join Bard talks.