Google Urges Employees to Rewrite Bard’s Terrible Responses

Google executives are aware that their company’s artificial intelligence search tool, Bard, isn’t always reliable in its responses. The onus to correct the incorrect responses rests, at least in part, with the workforce.

Google’s VP of Search, Prabhakar Raghavan, sent an email to employees on Wednesday asking for their assistance in vetting the accuracy of answers provided by Google’s newest ChatGPT opponent. CNBC saw an internal testing email from Bard with a link to a “do’s and don’ts” page with suggestions for improving test results. Employees are urged to revise their responses to questions they have a firm grasp on.

“Bard learns best by example, so taking the time to rewrite a response thoughtfully will go a long way in helping us to improve the mode,” the document says.

CNBC also stated that on Wednesday, CEO Sundar Pichai asked employees to devote two to four hours to Bard because “this will be a long journey for everyone, across the field.”

Google Urges Employees to Rewrite Bard's Terrible Responses
Google Urges Employees to Rewrite Bard’s Terrible Responses

If you’re interested in learning more about Google, you may click the links below to read about some of the news we’ve already covered:

Raghavan shared the same view:

“This is exciting technology but still in its early days,” Raghavan wrote. “We feel a great responsibility to get it right, and your participation in the dogfood will help accelerate the model’s training and test its load capacity (Not to mention, trying out Bard is actually quite fun!).” as per reported by CNBC.

The stock price of Google fell by over 9 percent after the company announced its chat technology last week, due to a series of mistakes around the unveiling. Internally, employees blasted Pichai for the blunders, using words like “rushed,” “botched,” and “comically short-sighted” to describe the deployment.

Also Read:  Kayla Durler Missing Update 2024: Is She Found Yet?

Executives are relying on human expertise to fix AI’s faults. First and foremost in Google’s list of do’s and don’ts is advice on things to think about “before teaching Bard.”

Google encourages its staff to keep their comments “polite, informal, and accessible,” or “dos.” Writing in the first person and keeping an unopinionated, neutral tone are two further requirements.

For don’ts, employees are told not to stereotype and to “avoid making presumptions based on race, nationality, gender, age, religion, s*xual orientation, political ideology, location, or similar categories.”

Also, “don’t describe Bard as a person, imply emotion, or claim to have human-like experiences,” the document says. And then Google advises, “keep it safe,” before instructing its staff to “thumbs down” responses that provide “legal, medical, financial advice” or are otherwise hostile or abusive. “Don’t try to re-write it; our team will take it from there,” the document says.

Raghavan has stated that those who participate in testing Bard and providing feedback would be awarded a “Moma badge” that will be displayed on their personal accounts within the company. Raghavan is in charge of the Knowledge and Information group, therefore Google will reportedly hold a listening session with the top 10 rewrite authors from that team.

Raghavan and the team behind Bard will be able to “share their feedback live” with the community. “A wholehearted thank you to the teams working hard on this behind the scenes,” Raghavan wrote.