Google Executives understand that the company’s Bard artificial intelligence search engine isn’t always accurate in how it responds to queries. At least some responsibility rests with the staff member for correcting incorrect answers.
Prabhakar Raghavan, Google’s vice president of search, asked employees in an email on Wednesday to help the company ensure that its new ChatGPT competitor gets the right answers. The email seen by CNBC included a link to a do’s and don’t page with instructions on how employees should correct responses as they internally audit Bard.
Employees are encouraged to rewrite answers on topics they understand well.
“Bard learns best by example, so taking the time to carefully rewrite answers will help us improve the regime in the long run,” the document reads.
Also on Wednesday, as CNBC previously reportedPichai asked employees to spend two to four hours on Bard, admitting that “it’s going to be a long journey for everyone, across the field.”
Raghavan echoed that sentiment.
“This is exciting technology but still in its early days,” Raghavan writes. “We feel a great responsibility to get it right, and your participation in dogfood will help speed up the model’s training and load testing (Not to mention, trying out the Bard is actually quite fun). !).”
Google disclosure its conversation technology last week, but a bunch of missteps around the announcement push the stock price down almost 9%. Workers criticize Pichai for the risks, describing the rollout internally as “hasty”, “unsuccessful” and “funny short-sighted”.
To try to correct AI’s mistakes, company leaders are relying on human knowledge. At the top of the do’s and don’ts, Google provides guidance on what to consider “before teaching the Bard”.
As a rule, Google instructs employees to keep answers “polite, casual, and approachable.” It also says they should be “in the first person” and maintain a “neutral, neutral tone.”
As for the do’s, employees are asked not to stereotype and “avoid making assumptions based on race, nationality, gender, age, religion, sexual orientation, major ideology value, location or similar categories.”
Additionally, “do not depict Bard as a person, imply emotions, or claim to have had human-like experiences,” the document reads.
Google then says “stay safe” and instructs employees to “disagree” with responses that offer “legal, medical, financial advice” or are hateful and abusive.
“Don’t try to rewrite it; our team will take it from there,” the document says.
To encourage people in his organization to check out Bard and provide feedback, Raghavan said contributors will earn a “Moma badge,” which appears on internal employee profiles. He said Google will invite the top 10 rewrite contributors from the Knowledge and Information organization, which is overseen by Raghavan, to a listening session. There, they can “share their feedback directly” with Raghavan and the people working on the Bard.
Raghavan wrote: “A big thank you to the teams for their hard work behind the scenes.
Google did not immediately respond to a request for comment.
