NY.- Google employees tasked with testing their employer’s AI chatbot, Bard, weren’t too pleased with what they found, Bloomberg reported.
After testing the bot, one employee allegedly called Bard “shameful” and another called him a “pathological liar,” according to screenshots of internal discussions.
Others found that the AI chatbot could be potentially dangerous. An employee said Bard was asked for advice on how to land a plane, and the advice he gave would lead to an accident, he said.
Another reportedly said that Bard generated diving tips “that would likely result in serious injury or death.”
One employee even begged Google to think twice before launching its chatbot, Bloomberg reported.
“Bard is worse than useless – please don’t launch him,” the employee wrote in a February internal message that was viewed by nearly 7,000 people.
Employees tasked with evaluating the safety and ethics of new Google products told Bloomberg that their employer discouraged them from trying to delay the development of its artificial intelligence technology, which they said lowered their morale.
They added that Google is de-prioritizing AI ethics to release a chatbot that can beat OpenAI’s ChatGPT as quickly as possible.
When Bloomberg reached out to Google for comment, Brian Gabriel, a spokesman said, “We continue to invest in the teams working to apply our AI principles to our technology.”
Despite employee concerns, Google decided to reveal Bard to the public during a disastrous demo in February that showed the chatbot making a factual error. A couple of days later, John Hennessy, the chairman of Alphabet, said Google wasn’t “really ready” to launch its AI. Google employees called the announcement “rushed” and “botched.”
The criticisms have only continued. A month after the demo, Google released a beta version of Bard. Like the employees, Early Access users weren’t impressed.
Users alleged that Bard spread misinformation, plagiarized articles and got basic math wrong, according to Gizmodo.
