Google’s brand-new AI-powered chatbot Bard still needs to pass the first test after providing a false response at a launch demonstration this week. Google uses the artificial intelligence (AI) search assistant that was announced on Tuesday to produce text summaries of search results. To promote the new feature launch, Google provided an animated image of Google Bard in use, but it provided an incorrect response. The lie will cause more people to doubt the veracity of search results and AI-generated answers to queries.
A user enters the search term “what discoveries from the James Webb Space Telescope can I tell my 9-year-old about” in an animated GIF demonstrating how Bard operates. The NASA telescope went into service in December 2021, and since then, researchers have found a number of new planets outside our solar system. “JWST took the very first pictures of a planet outside of our own solar system” is one of the responses Bard came up with. However, this is untrue.
When contacted about the bizarre story of Bard, Google did not respond right away
The Very Large Telescope array in Chile managed to snap the first image of an exoplanet, a planet outside our solar system, in 2004. Its name is 2M1207b, and it orbits a star 170 light-years from Earth. It is nearly five times as big as Jupiter. Google did not immediately answer requests for comment on Bard’s wild narrative.
There have been concerns about artificial intelligence systems producing errors that are difficult for humans to see. The developers of ChatGPT’s competing chatbot, OpenAI, have acknowledged their technology’s limitations and admit that it occasionally generates plausible-sounding but inaccurate or illogical responses to questions posed by people.
Company trained Bard using comparable data sources
The Telegraph quoted Questionmark inventor John Kleeman as saying in December that the technology was “fantastically impressive” and “really showing us how AI is going to impact the world.” The complete English-language material of Wikipedia, eight years’ worth of web pages scraped from the public internet, and scans of English-language books are just a few examples of the training data used by OpenAI for ChatGPT.
Although Google has yet to state how Bard was taught to produce its replies and summaries of search results, it is believed that the corporation used comparable data sources to train Bard.
A Google spokesman said: “This highlights the importance of a rigorous testing process, which we’re kicking off this week with our Trusted Tester program. We’ll combine external feedback with our own internal testing to make sure Bard’s responses meet a high bar for quality, safety, and groundedness in real-world information.”