What you need to know
- The success of Microsoft and OpenAI have pushed Google to compete within the AI space on an accelerated timeline.
- Google recently launched its own AI chatbot, Bard, in preview.
- Employees warned Google of Bard being “worse than useless” and giving advice that could “result in serious injury or death,” according to a report by Bloomberg.
Google’s chatbot Bard is once again in the headlines for a reason the tech giant likely isn’t happy about. According to a report by Bloomberg, Google employees warned the tech giant that Bard was a “pathological liar.” More than just sharing false information, that could be relatively harmless, Bard reportedly gave advice which “would likely result in serious injury or death” for activities such as scuba diving and landing a plane.
Back in February 2023, a Google employee reportedly sent an internal message that read, “Bard is worse than useless: please do not launch.” Almost 7,000 people were said to have seen that message, many of which agreed with its sentiment.
Jen Gennai, the AI governance lead at Google, overruled a risk assessment from her team in March, according to Bloomberg. Bard launched to the public in preview shortly after.
According to the report, Gennai’s decision was made in conjunction with a group of senior leaders that “determined it was appropriate to move forward for a limited experimental launch with continuing pre-training, enhanced guardrails, and appropriate disclaimers.”
Google’s quickened pace in the AI race is due, at least in part, to OpenAI releasing ChatGPT and Microsoft integrating AI into services such as Bing. Last December, Google’s senior leadership reportedly sent out a “code red” and accelerated its efforts in artificial intelligence. That change of course required Google to make some compromises when it came to ethics.
Gennai reportedly said that when it comes to “fairness,” Google “might be at 80, 85 percent, or something.” Those figures refer to a threshold that Google products have to hit before being released. Some products, such as those related to child safety, have to hit a 100% threshold.
Google said the following to Bloomberg regarding Bard:
“We are continuing to invest in the teams that work on applying our AI Principles to our technology.”
