Google and other major tech companies this week have been showcasing how conversational chatbots can help improve internet search.
In one instance, however, Google may have inadvertently showcased the technology’s shortcomings.
Built on Google’s own large language models (LLMs), Bard is designed to give users conversational answers to relatively complex questions.
To demonstrate this, the GIF shows Bard answering the question, “What new discoveries from the James Webb Space Telescope can I tell my 9 year old about?” Bard’s answers, at first glance, appear useful. However, observers soon took note that the chatbot was delivering inaccurate information.
“JWST took the very first pictures of a planet outside of our own solar system,” Bard said in its reply.
In fact, the first photo of an exoplanet was taken in 2004 by the European Southern Observatory’s VLT (Very Large Telescope).
While Google’s tweet was still online as of Wednesday morning (and viewed by more than 1 million Twitter users), Google acknowledged the error in a statement provided to ZDNET.
“This highlights the importance of a rigorous testing process, something that we’re kicking off this week with our Trusted Tester program,” a Google spokesperson said in a statement. “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.”
While Google has said it plans to roll out Bard to the general public in the next couple of weeks, the chatbot is initially undergoing testing by a select number of users.
Meanwhile, Microsoft on Tuesday unveiled a new version of the Bing search engine running on a next-generation OpenAI large language model, making it “more powerful than ChatGPT” — a conversational chatbot that is quickly gaining in popularity. OpenAI, the makers of ChatGPT, also recently introduced its own premium version of the chatbot, ChatGPT Plus.