Google relies on human employees to improve Bard chatbot’s responses
In a video ad Google posted on Twitter, its yet-to-be-launched AI chatboard Bard confidently spouted misinformation about the James Webb Space Telescope. “JWST took the very first pictures of a planet outside of our own solar system,” the chatbot replied, which is patently false. (It was the European Southern Observatory’s Very Large Telescope that captured images of exoplanets for the first time.) Now, the tech giant is looking to improve Bard’s accuracy, and according to CNBC, it’s asking employees for help.
Google’s VP for search, Prabhakar Raghavan, reportedly sent an email to staff members, asking them to rewrite Bard responses on topics they know well. The chatbot “learns best by example,” Raghavan said, and training it with factual answers will help improve its accuracy. Raghavan also included a list of “dos” and “don’ts” when it comes to fixing Bard’s responses, based on the email seen by CNBC.
Responses should be in first person POV, should be unopinionated and neutral, and they should have a polite, casual and approachable tone. Employees are also instructed to “avoid making presumptions based on race, nationality, gender, age, religion, sexual orientation, political ideology, location, or similar categories.” They’re asked not to describe Bard as a person, imply that it has emotions or claim that it has human-like experiences. Plus, they’re instructed to thumbs down any responses the chatbot might give containing “legal, medical, financial advice” or are hateful and abusive.
Raghavan’s memo came after Google CEO Sundar Pichai emailed employees, asking them to spend a few hours each week testing the AI chatbot. Google employees reportedly criticized Pichai for a “rushed” and “botched” Bard rollout. The CEO is now giving staffers the chance to “help shape [the chatbot] and contribute” by testing the company’s new product. He also reminded everyone that some of Google’s “most successful products were not first to market” and that they “gained momentum because they solved important user needs and were built on deep technical insights.”
People have been anticipating Google’s response to ChatGPT ever since the OpenAI chatbot arrived late last year. The Microsoft-backed technology has gained tremendous popularity over the past few months, enough to rattle Alphabet and its investors. Google tried to assuage investors’ concerns during its quarterly earnings call in early February by talking about its own chatbot and by touching on its work developing an AI-powered Search to compete with the next-gen Bing.
All products recommended by Engadget are selected by our editorial team, independent of our parent company. Some of our stories include affiliate links. If you buy something through one of these links, we may earn an affiliate commission. All prices are correct at the time of publishing.
Comments are closed, but trackbacks and pingbacks are open.