Mark Cuban says he’s a fan of artificial intelligence — so much so that it’s become a regular part of his day-to-day routine. The billionaire entrepreneur and startup investor uses AI for “everything,” he told the “High Performance” podcast in a June 30 episode. That includes writing code for software development, a skill Cuban hadn’t previously used in years, he added. “It’s insane how much I use it right now. I downloaded this app, Replit, and you just type in … I want to be able to compare pharmacy costs between my company and these other companies, and every time the price changes send me alerts,” said Cuban, 66.

“Within a few minutes, it had the first pass of the software. Then, I just ran it multiple times and gave it new ideas and things I wanted.”

He also uses AI to make text-to-video content for the NBA’s Dallas Mavericks — he’s a minority owner of the team — and to monitor his health, he said. “I recently had this thing called afib [atrial fibrillation] and I had an ablation, so I had to track all the things I was doing, the drugs I was taking. And in the past, you might have downloaded an app and marked things down,” said Cuban. “I just [went to ChatGPT] and I was like … I’m going to tell you when I take my medicine and when I do my workout, and I want you to record it all. If there’s something there that I write down that you think is not right or I should be concerned about, let me know.”

The technology isn’t perfect, Cuban said: “An advanced programmer could do a better job.” DON’T MISS: A step-by-step guide to buying your first home—and avoiding costly mistakes But if you’re not an advanced programmer, you can at least get comfortable with prompt engineering, the process by which you train an AI to give you the output you desire, noted Cuban. Just make sure you fact check anything it tells you, he said: AI chatbots can hallucinate, generating fabricated information and presenting it as fact. “Chatbots were generally bad at declining to answer questions they couldn’t answer correctly, offering incorrect or speculative answers instead,” found researchers at Columbia University’s Tow Center for Digital Journalism, in a report published on March 25. Paid chatbots typically presented their incorrect answers more confidently than free ones, and some of the chatbots generated fake links when asked to “accurately retrieve and cite news content,” the researchers found.

“You’ve got to be careful … It’s like talking to a friend who you think knows a lot about something,” said Cuban. “You’ve still got to be careful and talk to an expert.” Some other tech billionaires — including multiple with vested interests in promoting AI‘s use — also say they use the technology regularly. OpenAI CEO Sam Altman uses the tech to process emails and summarize documents, he told Adam Grant’s “ReThinking” podcast in January.

Microsoft CEO Satya Nadella, unsurprisingly, uses Outlook’s AI features to organize and arrange his inbox, he said at the Fast Company Innovation Festival 2024. Nvidia CEO Jensen Huang, whose company manufactures computer chips used for training and developing AI models, uses chatbots to help him write drafts, he said at a Wired event in December. “I give it a basic outline, give it some PDFs of my previous talks, and I get it to write my first draft,” said Huang. “It’s really fantastic.”

CNBC – via Reuters Connect