Sundar Pichai Opens Up About Google Bard's Trippy Troubles: 'No One In The Field Has Yet Solved Hallucination Problems'

Alphabet Inc. GOOG GOOGL CEO Sundar Pichai acknowledged that artificial intelligence or AI technology generally is still struggling with “hallucination problems” with no clear explanation. 

What Happened: Since OpenAI released its revolutionary chatbot technology called chatGPT for public testing, the term “hallucination” has become a beacon of concern for tech experts and netizens alike. 

During a CBS News’ 60 Minutes interview, Pichai acknowledged AI “hallucination problems,” saying, “No one in the field has yet solved the hallucination problems. All models do have this as an issue.” 

See Also: ‘Un-Googley’: Google Employees Ridicule CEO Sundar Pichai For Rushing Bard AI Chatbot Announcement

For the unversed, in the AI ecosystem, the term hallucination refers to AI giving confident answers that don’t seem to be justified by its training data. 

Pichai took to Twitter to share the interview in which he admitted that solving this problem is a “matter of intense debate.”

He also explained that to “help cure hallucinations,” Google Bard features a “Google it” button that leads users to the old-fashioned search. 

Pichai also talked about the “most mysterious” AI issue called “emergent properties.” 

“Some AI systems are teaching themselves skills that they weren’t expected to have. How this happens is not well understood,” he stated. 

Why It’s Important: In February earlier this year, Google launched its chatbot, Bard. However, the demonstration didn’t go as planned, as the AI chatbot gave an incorrect answer to a question that sparked massive backlash and debate over its accuracy. 

Previously it was reported that for the first time in 25 years, the tech giant’s business is facing a threat from age-old rival Microsoft Corporation MSFT as the latter is thriving on the success of its Bing AI, powered by the same OpenAI technology that works behind chatGPT. 

Check out more of Benzinga’s Consumer Tech coverage by following this link.

Read Next: Elon Musk Is Worried About AI Posing A Threat To Humanity, But Prepares To Launch OpenAI Rival

Photo by World Economic Forum on Flickr