Google cautions against ‘hallucinating’ AI chatbots: report

“This kind of artificial intelligence we’re talking about right now can sometimes lead to something we call hallucination,” Prabhakar Raghavan, senior vice-president at Google and head of Google Search, told Germany’s Welt am Sonntag newspaper.

“This then expresses itself in such a way that a machine provides a convincing but completely made-up answer,” Raghavan said in comments published in German. One of the fundamental tasks, he added, was keeping this to a minimum.

Google has been on the back foot after OpenAI, a startup Microsoft is backing with around $10 billion, in November introduced ChatGPT, which has since wowed users with its strikingly human-like responses to user queries.

Alphabet Inc introduced Bard, its own chatbot, earlier this week, but the software shared inaccurate information in a promotional video in a gaffe that cost the company $100 billion in market value on Wednesday.

Alphabet, which is still conducting user testing on Bard, has not yet indicated when the app could go public.

Discover the stories of your interest

“We obviously feel the urgency, but we also feel the great responsibility,” Raghavan said. “We certainly don’t want to mislead the public.” A costly error

The company posted a GIF of its new AI Bard in action which showed the chatbot giving a factually inaccurate response to a prompt.

In the GIF, the chatbot is prompted, “What new discoveries from the James Webb Space Telescope (JWST) can I tell my 9-year old about?”

Bard responded with a number of answers, including one suggesting the JWST was used to take the very first pictures of a planet outside the Earth’s solar system, or exoplanets.

This is where it went wrong as the first pictures of exoplanets were, however, taken by the European Southern Observatory’s Very Large Telescope (VLT) in 2004, as confirmed by NASA.

Alphabet is coming off a disappointing fourth quarter results as advertisers cut online spends. The search and advertising giant is moving quickly to keep pace with OpenAI and rivals, reportedly bringing in founders Sergey Brin and Larry Page to accelerate its efforts.

Google CEO Sundar Pichai on Monday, announced in a blog post that the company was opening Bard to a group of trusted testers, ahead of making it more widely available to the public in coming weeks.

“Two years ago we unveiled next-generation language and conversation capabilities powered by our Language Model for Dialogue Applications (or LaMDA for short). We’ve been working on an experimental conversational AI service, powered by LaMDA, that we’re calling Bard,” read the blog post.

Also read | ChatGPT, Bard & Ernie: The three musketeers of AI