Recently, ChatGPT, a generative AI application launched by OpenAI, has exploded, even described by Bill & middot; Gates as no less important than the birth of the Internet, leading a number of technology giants to announce that they will launch similar AI programs.

Among them, Google, which is the fastest, has shown the Bard of ChatGPT not long ago, and has access to a lot of promotion resources.

But embarrassingly, Bard gave the wrong answer when answering the query.As a result, Google's parent company Alphabet's share price plummeted 8%, reducing its market capitalization by more than $100 billion.

In response, Prabhakar Raghavan, senior vice president of Google and head of Google Search, said& ldquo; the kind of artificial intelligence we are talking about now sometimes leads to what we call an illusion in which the machine gives a seemingly convincing but completely fabricated answer. & rdquo

To put it simply, he believes that in addition to Bard, ChatGPT or other similar chat AI also have serious nonsense and give the wrong answer.

In fact, there are indeed many users who sayWhen I use all kinds of chat AI, I have the wrong answer or give the wrong answer.

Keeping this kind of situation to a minimum and reducing its possible impact is still one of the basic tasks of the artificial intelligence industry.