Google CEO says ‘things will go wrong’ with chatbot rollout

Google is slowly releasing its chatbot Bard to the public, but the company’s CEO just warned employees that although thousands of people had already tested the A.I., it might still have a lot of problems. 

“Even after all this progress, we’re still in the early stages of a long A.l. journey,” Pichai wrote in the Tuesday memo to employees, according to CNBC. “As more people start to use Bard and test its capabilities, they’ll surprise us. Things will go wrong.” 

Until now, Google has kept interactions with Bard limited to hand-picked testers, including about 80,000 Google employees, according to Pichai. But on Tuesday the company announced it was rolling out beta testing to thousands of users in the U.S. and U.K. who join the waitlist, with languages other than English to be added over time.  

“We’ve taken a responsible approach to development, including inviting 10,000 trusted testers from a variety of backgrounds and perspectives, and we’ll continue to welcome all the feedback that’s about to come our way. We will learn from it and keep iterating and improving,” Pichai wrote to employees. 

It appears that the more widely-available Bard has already run into a hiccup. Hong Kong-based security researcher Jane Manchun Wong wrote on Twitter Tuesday that she asked Bard whose side it would pick in the antitrust lawsuit against Google filed by the U.S. Department of Justice, and it promptly sided with the government.

“Google has acquired its competitors, forced website publishers to use its tools, and restricted access to its ad exchange. These actions have harmed consumers and stifled innovation in the digital advertising market,” Bard responded. “I hope that the court will find in favor of the Justice Department and order Google to take steps to break up its monopoly.” 

A Google spokesperson told Fortune that Bard was still in its early stages and its purpose is to give users multiple viewpoints.

“Since LLMs like Bard train on publicly available content, they can reflect positive or negative views of specific politicians, celebrities or other public figures, or even incorporate views on certain sides of controversial social or political issues into their responses,” the spokesperson said. “As we’ve said, Bard can sometimes give inaccurate or inappropriate information that doesn’t represent Google’s views and Bard should not respond in a way that endorses a particular viewpoint on subjective topics.”

Google was caught-flat footed by the release of Open AI’s ChatGPT, which has taken the internet by storm over the last few months with a major investment in Microsoft, which is incorporating it into its Bing search engine. Google has poured money into A.I. research for years, but was forced by ChatGPT to bring its product to the market quickly.

“There are other companies taking shots at us saying that we are late to market. We’ve missed the A.I. boat,” Google Cloud CEO Thomas Kurien reportedly said in a leaked audio recording about the company’s nay-sayers. “I would tell you this is the very first minute of a new game, and the game is never done in the first minute.”

But even Google’s soft rollout has not been without some big mistakes. In a promotional video released by the company, Bard inaccurately answered a question about the James Webb Space Telescope, and the mistake cost the company $100 million in market value. Other chatbots including ChatGPT have also answered questions inaccurately, and displayed other disturbing behavior, in one instance suggesting to a reporter that he was unhappy in his marriage and should leave his wife.  

Bard opened up to limited public users over a month after Microsoft’s Bing A.I. bot did and about four months after the OpenAI-owned ChatGPT frenzy started. But Google was the first to announce using A.I. in its office tools like Google Docs and Gmail. Microsoft said it would introduce versions of its bot on its Office products including in powerpoint presentations and word documents the same week. OpenAI, in the meantime, has launched a version of ChatGPT for business and a new-and-improved version of its chatbot technology, GPT-4 that is used to power ChatGPT and Microsoft’s Bing searches.

Subscribe to Well Adjusted, our newsletter full of simple strategies to work smarter and live better, from the Fortune Well team. Sign up today.

Source link