Clicky

Home » Latest News » Google Bard Demo: What went wrong with Google’s AI Chatbot

Google Bard Demo: What went wrong with Google’s AI Chatbot

By

Andrew NG

| Updated on:

Google’s Bard AI was one of the most anticipated AI chatbots in the market. However, the very first demo of this AI chatbot turned into a failure when it made a factual error and delivered an incorrect answer to one of the questions.  

Google’s parent company even faced a decline of $100 billion in market value after the occurrence of a factual error. Stating their concern that Microsoft might leave Google behind in the AI market.

In this post, we will look into Google Bard Demo and what went wrong with Google’s AI chatbot. In addition, what happened after Google’s Bard demo failed?  

About Google Bard Demo

Google’s much-awaited AI Chatbot Bard released a promotional video last week where they shared a demonstration of Bard AI. However, this AI chatbot made a factual error in its very first demo video. Where Bard delivered an incorrect answer to the question: “What new discoveries from James Webb Space Telescope (JWST) can I tell my nine year old?”. 

It stated JWST took the very first picture of a planet outside of our solar system. According to NASA, it was “European Southern Observatory’s giant telescopes that took the very first photo. 

This resulted in quite a storm on the internet by various astrophysicists for incorrect information. Google’s parent company even saw a drop of $100 billion in its market value after the factual error.  

When did Google show Google Bard Ai Demo?

Google revealed its new AI chatbot “Bard AI” on 8th February 2023, in Paris, France. Google narrated its AI conversation technology LaMDA. However, they gave only a few demonstrations of Bard at work. 

During this presentation, they covered augmented reality along with Google Maps and linked it to AI-powered. Displaying a demo of a person trying to find a coffee shop through maps. The presentation was more about the promise of innovation rather than witnessing Bard AI in action. 

Hours after the live stream event in Paris, Google issued an advertisement demonstration of their new AI chatbot Bard. The ad showcased Bard responding to the question, What new discoveries from James Webb Space Telescope, can I tell my nine year old?. 

To which Bard responded with three bullet points. However, while responding, it turned out the third point delivered by Bard stating “JWST took the very first picture of a planet outside of our solar system” was false. 

According to NASA, the first-ever picture of a planet taken from outside of our solar system was taken by European Southern Observatory’s giant telescopes. This resulted in people questioning the accuracy of Bard’s information. 

It also caused a collapse in shares of Google’s parent company “Alphabet” which continued for around two days. Around 7.7% of shares were wiped out causing a loss of $100 billion off its market value. 

What was the error in Google Bard Demo?

Google Bard demo faced a Factual error, as it was criticized by experts and astrophysicists for providing incorrect information. In the promotional video of Google Bard, the new AI chatbot was seen answering the question: “What new discoveries from James Webb Space Telescope, can I tell my nine year old?”. 

However, Bard delivered incorrect information by stating the wrong name of the telescope that took the first image of an Exoplanet. Several astrophysicists tweeted regarding the false information. Grant Tremblay tweeted “JWST didn’t take the first-ever image of a planet outside our solar system”. 

Bruce Macintosh (Director of the University of California Observations) stated in a tweet: “As somebody who has been imaging an exoplanet for about fourteen years before JWST was developed. Feels like you should have looked for a better example”. 

What happened after Google Bard Demo went wrong

After Google’s AI-powered chatbot Bard’s Demo failed, Google’s parent company “Alphabet” witnessed a drop of $100 billion in their market value. This fuelled the concern among a few shareholders regarding Google’s parent company failing behind Microsoft in the AI race.   

However, a statement from Google’s spokesperson did come along after the demo failed to state, “The factual error raised the importance of rigorous testing process, something they are kicking off from this week onwards along with our Trusted tester program. 

The spokesperson also claimed “We are combining the external feedback of our internal testing to certify Bard’s responses meet a high bar of safety, quality, and groundedness in real-world information. 

Related Articles:

Leave a Comment