We already know that Google would "appreciate" it if its employees take quite a bit of their work time to test the upcoming Bard chatbot AI feature. Now, another company wide email is giving some specifics to those workers on how to write answers for Bard.
CNBC reports that in the email from Prabhakar Raghavan, Google’s vice president for Search, he asks employees to use their own knowledge of topics that they are interested in to rewrite Bard's answers to those questions. He stated, "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."
In terms of how to rewrite an answer, Raghavan says that they should be using a first-person viewpoint, and have a neutral tone. The answers should also be "polite, casual and approachable." Answers should not suggest that Bard is actually a person, and they should avoid any emotional or human-like responses.
Employees are also asked that their Bard answer suggestions not make any presumptions on a person's race, religion, gender or other similar kinds of classifications. Any answers that offer advice on financial, legal, or medical questions are outright banned.
If a Google employee does make contributions to Bard testing, the email states that they will get a "Moma badge" on their company profile. The workers with the best answers for the chatbot will get face-to-face meetings with the Bard team. So far, Google has yet to reveal when Bard will be made available for public testing.
6 Comments - Add comment