Shares of Google’s parent company Alphabet plunged yesterday after its new AI chatbot incorrectly answered a question during a promotional video.
The Google owner saw shares fall nearly eight per cent on Wednesday, knocking $100bn (£82bn) off its value.
This comes after Alphabet announced that Google will be integrating its own version of an AI chatbot similar to ChatGPT called Bard.
This was followed by a clip released of a user asking Bard “What new discoveries from the James Webb Space Telescope (JWST) can I tell my nine-year-old about?”.
In his answer, Bard included the statement that JWST took the first picture of a planet outside Earth’s solar system.
Experts pointed out an embarrassing flaw in the bot’s response, with astrophysicist Grant Tremblay posting on Twitter: “Shouldn’t be a good one, really~ blow, and I’m sure Bard would be impressive, but for the record: JWST Took “the first image of a planet outside our solar system”
He corrected Bard, saying that the first image was actually taken by “Chauvin et al”. (2004) using adaptive optics with VLT/NACO”.
This error highlights a major concern with AI, which is that, while its answers may seem correct, they are not always factually correct and do not cite their sources.
Google introduced Bard on Monday, although it did not specify an exact date for when the platform would be publicly available.
Alphabet CEO Sundar Pichai commented on the reveal: “Bard seeks to combine the breadth of the world’s knowledge with the power, intelligence and creativity of our large language model. It harnesses information from across the web to provide fresh, high-quality responses. receives”.
The company recently announced that it is investing $300 million in Anthropic, an AI startup by former OpenAI members.
Hours later a live-streamed event took place in Paris, where the tech giant was trying to promote its new chatbot named Bard.