ChatGPT, A.I. tech will make internet misinformation worse

Mark Cuban may be entertained by chatbots like Microsoft-backed ChatGPT and Google’s upcoming Bard — but he isn’t ready to trust them.

Online misinformation “is only going to get worse” as artificial intelligence platforms evolve and spread, the billionaire tech entrepreneur and investor said on a recent episode of comedian Jon Stewart’s podcast, “The Problem with Jon Stewart.”

Right now, misinformation tends to spread through social media platforms like Facebook or Twitter — and that’s with some semblance of human guardrails in place, Cuban said. But with ChatGPT and other similar platforms, the machines are in control.

“Once these things start taking on a life of their own … it will be difficult for us to define why and how the machine makes the decisions it makes, and who controls the machine,” Cuban said.

Hundreds of millions of users have tried ChatGPT to write poems, offer advice and recite recipes since the platform launched in November. But so far, the technology isn’t showing itself to be smarter than the average human.

Posting the chatbot’s simplistic errors is a popular social media trend. At times, ChatGPT incorrectly answers math problems, refuses to answer basic riddles and even “hallucinates”— or completely makes up historical figures, events and other details that seem like facts.

ChatGPT can also contradict itself, sometimes providing different answers when repeatedly asked the same question.

Similarly, shares of Google’s parent company Alphabet dropped more than 9% this week after Bard incorrectly answered a question about NASA’s James Webb Space Telescope in one of Google’s first ads for the AI platform.

A raft of Google employees have blamed CEO Sundar Pichai for Bard’s “rushed, botched” release, with the company feeling pressured to compete with ChatGPT, CNBC reported on Friday.

“Rushing Bard to market in a panic validated the market’s fear about us,” read one post on an internal Google forum reviewed by CNBC, alongside a photo of a face-palming bird.

The errors show that the technology is still in infantile stages. That’s a problem, especially for large swaths of people who don’t always fact check claims they see on the internet, Cuban said.

“Our generation, Gen X and older, doesn’t get it,” Cuban said. “Gen Z and younger, they’re not only native to it, they know how to block things out … They’re more in tune to all these issues.”  

Microsoft, for its part, acknowledges that the technology behind ChatGPT isn’t perfect — even as it plans to incorporate it into an upcoming version of its search engine, Bing.

“Bing will sometimes misrepresent the information it finds, and you may see responses that sound convincing but are incomplete, inaccurate, or inappropriate,” the company’s recently updated FAQ page says.

In the short term, that could be a problem — a concern Cuban shares with fellow tech billionaire Steve Wozniak. But other industry luminaries have expressed excitement about the technology’s longer-term possibilities.

Microsoft co-founder Bill Gates, for example, thinks platforms like ChatGPT represent a burgeoning technological revolution that’ll make a “huge impact” on health care and education, he told German-language business newspaper Handelsblatt’s “Disrupt” podcast on Thursday.

“Today, they require too much computation, they’re not always accurate … But even this week, you’ll have announcements from Microsoft and Google, where they’re competing to lead in this space,” Gates said. “The progress over the next couple of years to make these things even better will be profound.”

Get CNBC’s free Warren Buffett Guide to Investing, which distills the billionaire’s No. 1 best piece of advice for regular investors, do’s and don’ts, and three key investing principles into a clear and simple guidebook.

Sign up now: Get smarter about your money and career with our weekly newsletter