The company is seeking feedback from users to compete with Microsoft’s ChatGPT in the rapidly developing artificial intelligence market. Previously, only approved testers could access Bard, which Google describes as an experiment in collaboration with generative AI. The technology relies on past data to create content, rather than identify it.
The release of ChatGPT, a chatbot from the Microsoft-backed startup OpenAI, last year, has triggered a race among technology companies to offer AI to more users. The aim is to transform the way people work and do business.
Last week, Google and Microsoft announced draft-writing technology for their word processors and other collaboration software. They also unveiled marketing-related tools for web developers to create their own AI-based applications.
When asked if the competitive dynamics were behind Bard’s rollout, Jack Krawczyk, a senior product director, said that Google was focusing on users. He stated that testers have used Bard to boost productivity, accelerate their ideas, and fuel their curiosity.
In a Reuters demonstration of the site, bard.google.com, Krawczyk showed how the program could generate blocks of text instantly, different from how ChatGPT produces answers word by word.
Bard also has a feature that displays three different versions or “drafts” of any given answer, among which users can toggle. Additionally, it has a button that says “Google it” to provide users with web results for their queries.
Unlike ChatGPT, Bard is not proficient in generating computer code, according to Google. The company also said that it had limited Bard’s memory of past exchanges in a chat and was not currently using it for advertising, which is core to Google’s business model.
Accuracy is still a concern. A Google pop-up notice warns during the demo that “Bard will not always get it right.” Last month, a promotional video showed the program answering a question incorrectly, resulting in a $100 billion loss in Alphabet’s market value. During the demonstration, Google highlighted a few mistakes made by Bard, including falsely claiming that ferns required bright, indirect light in response to one query.
In another instance, Bard produced nine paragraphs of text when asked for four. After that answer, Krawczyk clicked a thumbs-down button for feedback.
“We know the limitations of the technology, and so we want to be very deliberate at the pace at which we roll this out,” said Krawczyk. Google’s aim is to enhance Bard’s performance before making it widely available to the public. The company’s strategy is to focus on user feedback to improve the chatbot’s accuracy and capabilities.
In conclusion, Google has publicly released Bard, a generative AI chatbot, to compete with Microsoft’s ChatGPT. The program relies on past data to create content and is currently limited to English-language access. Bard’s rollout aims to gain feedback from users to enhance the chatbot’s accuracy and capabilities.
However, the technology still has limitations, and Google is being cautious about the pace at which it rolls out Bard to the public.