Google's much-hyped new AI chatbot tool Bard, which has not yet been released to the public, was criticized this week for producing an erroneous response during a demonstration.
In the Google-tweeted demonstration, a user asks Bard, "What discoveries from the James Webb Space Telescope can I tell my nine-year-old about?" Bard responds with several bullet points, including one that reads, "JWST took the very first pictures of a planet outside of our solar system."
NASA reports that the first photograph of an exoplanet, or any planet beyond our solar system, was captured by the Very Large Telescope of the European Southern Observatory about two decades ago, in 2004.
The incorrect response from Bard was first published by Reuters around midday, causing shares of Google's parent company, Alphabet, to fall as much as eight per cent, or approximately $140 billion.
Bard's error illustrates the difficulty Google faces as it attempts to integrate the same artificial intelligence technology that drives Microsoft's ChatGPT into its primary search engine.
To stay up with what some believe could be a significant shift in how people conduct internet searches prompted by conversational AI, Google now risks damaging its search engine's reputation for providing accurate results.
Bard, like ChatGPT, is built on a big language model trained on enormous web data troves to provide engaging responses to user inputs. Experts have long warned that these tools may facilitate the dissemination of false information.
To address this problem, Google stated that Bard would be made accessible to "trusted testers" this week, with intentions to release it to the whole public in the following weeks.
Regarding the factual error, a Google representative told CNN in a statement, "This highlights the importance of a rigorous testing process, something that we're kicking off this week with our Trusted Tester program,"
"We'll combine external feedback with our internal testing to ensure Bard's responses meet a high bar for quality, safety and groundedness in real-world information."
Google introduced Bard earlier this week in what appears to be an attempt to compete with the viral success of ChatGPT, which has been used to generate essays, song lyrics, and answers to questions previously searched for on Google. According to reports, ChatGPT's stratospheric surge in popularity has forced Google's executives to declare a "code red" for their search product.
Google's Paris headquarters hosted an event on Wednesday where the tech giant unveiled intentions to use AI technology to alter how individuals search for information online fundamentally. The Google event occurred one day after Microsoft announced a revised version of Bing powered by a more powerful version of the artificial intelligence utilized by ChatGPT. (Microsoft is investing billions of dollars in OpenAI, which developed ChatGPT.)
In the presentation, a Google official hinted at ambitions to leverage this technology to provide more complex and conversational solutions to searches, such as bulleted lists of the ideal times of year to observe particular constellations and pros and drawbacks of purchasing an electric vehicle.
The executive stated that artificial intelligence (AI) technology would pave the way for the "next frontier of our information products."
