Google’s wannabe rival to ChatGPT is off to a shaky start after its launch video featured a glaring error about the JWST and exoplanets. As a result of the blunder, shares in parent company Alphabet plunged by around $100 billion on Wednesday.
In a promo video posted on Wednesday, Googles's new artificial intelligence (AI) chatbot, Bard, was asked to describe the discoveries made by JWST to a nine-year-old child. It replied that it was the first telescope to ever take pictures of a planet outside of the Solar System.
Unfortunately, however, that is not true.
As pointed out by a number of astronomers on Twitter, the first image of an exoplanet was actually taken almost two decades ago, long before JWST was launched on Christmas Day 2021. Taken by the Very Large Telescope in 2004, the first exoplanet image shows a distant world orbiting the brown dwarf 2M1207 a mere 230 light-years away in the constellation of Hydra.
It appears that Google’s AI bot was getting confused with recent news headlines that explained how JWST had captured its first image of an exoplanet.
"Why didn't you fact check this example before sharing it?" Chris Harrison, an astronomer and fellow at Newcastle University in the UK, replied to the tweet.
It proved to be a costly mistake. Alphabet's shares reportedly dropped by 9 percent on Wednesday, wiping around $100 billion from the company's market value, Reuters reported.
ChatGPT is a free-to-use chatbot developed by OpenAI that can generate human-like replies to questions and demands. The technology is based on neural networks, which mimic the underlying architecture of the brain to process information and learn.
It’s pretty remarkable. Ask it, for example, to write a news article about exoplanets in the style of IFLScience and it will do so with ease (not that we’d ever do that, though some places have already). It’s so smart it was even capable of passing a US medical licensing exam.
There’s a lot of hype around ChatGPT with many claiming it has sparked an accessible AI revolution that will be as big as the advent of the personal computer or the Internet.
It also has the potential to challenge Google’s tech supremacy by posing “an existential threat” to its search model. The creator of Gmail recently warned that AI could totally upend Google within a few years by eliminating the need for a search results page, which is how the tech giant makes tons of money through ad revenue.
Google realizes this threat and has launched Bard to rival ChatGPT. By the looks of things, it still has a way to go. However, Google insists it’s all just part of the process.
"This highlights the importance of a rigorous testing process, something that we're kicking off this week with our Trusted Tester program," a Google spokesperson reportedly said in a statement.
"We'll combine external feedback with our own internal testing to make sure Bard's responses meet a high bar for quality, safety, and groundedness in real-world information."