How Is Mark Cuban Using AI Every Day in Business?

TL;DR

Mark Cuban uses AI daily to write software with Replit, generate text-to-video clips for the Mavericks, and track his medication and workouts through ChatGPT after an ablation for AFib. He argues AI will produce the world's first trillionaire, possibly one person working alone in a basement, because the technology changes daily and nobody has seen its best or craziest capabilities yet.

Transcript

when you see a thousand people doing the same thing why would you want to be the thousandth and first i can just go on All the [ __ ] I'm trying with with AI What would you like to say to the AI naysayers you have no say in the matter All this stuff is just changing every single day We haven't seen the best or the craziest of what it's going to be... Read More

Key Insights

• AI will create the world's first trillionaire, according to Cuban, because the technology changes every single day and someone will use it in a way nobody has considered. It could be one person working alone in a basement rather than a large company.
• Every major technology wave met the same resistance before becoming indispensable. People said they did not need PCs, questioned what the internet was for, dismissed phone cameras and Snapchat, then five years later could not imagine living without them. Cuban expects the same arc for AI.
• Replit let Cuban write working software despite not keeping his programming skills current. He typed a request for an app comparing pharmacy costs across companies with price-change alerts, got a first pass in minutes, then iterated by feeding it new ideas repeatedly.
• The software AI produced was not perfect, and Cuban says an advanced programmer could do a better job. The value was that a non-current coder could go from a typed description to running software through repeated passes rather than writing it himself.
• Text-to-video tools now turn a sentence into footage. Cuban describes asking for a person carrying a rugby ball who drop kicks it into a basketball basket and receiving roughly ten seconds of video in about a minute, which he uses for the Mavs and other businesses.
• ChatGPT can act as a health log. After an ablation for AFib, Cuban told it when he took medicine and worked out and asked it to flag anything concerning, which surfaced things he would not have known to look for.
• AI output should be challenged the way you would challenge a person. Cuban tells the model he does not think it is right and demands the evidence behind its conclusion, treating that as part of the process alongside consulting an actual expert.
• Doing something the way it has always been done is a red flag, in Cuban's view. When a thousand people are doing the same thing, there is no reason to be the thousand and first, and the world does not need another Uber or another DoorDash.

Questions & Answers

Q: Why does Mark Cuban think AI will create the first trillionaire?

Cuban points to the speed of change. The industry and the underlying technology are changing so rapidly that somebody will come up with a new way to use it that nobody has thought about before. He draws the parallel to the early days of PCs, networks, the internet, mobile phones, and 5G, where an innovative entrepreneur always built something bigger and better, but says AI dwarfs all of those. He adds that we have not seen the best or the craziest of what it will be able to do, and that the trillionaire could be just one person in a basement.

Q: What is Mark Cuban using AI for on a day-to-day basis?

Almost everything, in his words. He used the application Replit to write software despite not having kept his programming skills current, describing an app to compare pharmacy costs between his company and others with alerts on price changes. He uses text-to-video tools for the Mavericks and other businesses, generating short clips from a typed description. He also used ChatGPT to track his medication and workouts after a heart procedure, and uses the voice interface on an Android phone, where he says you get caught up talking to it like a person.

Q: How did Mark Cuban build software without current programming skills?

He downloaded an application called Replit and typed in a plain description of what he wanted: an app that compares pharmacy costs between his company and other companies, and sends him alerts every time the price changes. Within a few minutes it produced a first pass of the software. He then ran it multiple times, feeding it new ideas and additional things he wanted. Cuban is clear that the result was not perfect software and that an advanced programmer could do a better job, but he still calls the experience crazy given how little he wrote himself.

Q: How did Mark Cuban use ChatGPT to track his health?

After being diagnosed with AFib and having an ablation, Cuban needed to track his medications and activity. Rather than downloading an app and marking things down manually as he might have in the past, he told ChatGPT he would report when he took his medicine and when he worked out, and asked it to record everything and flag anything that looked wrong or worth being concerned about. He says it did exactly that. He adds a caution: it is like talking to a friend who you think knows a lot about a subject, so you still need to talk to an actual expert.

Q: Should you trust AI answers, according to Mark Cuban?

Not blindly. Cuban says that even though it is called artificial intelligence, that does not mean it will always be right, now or forever, so questioning it is part of the process. When he thinks an answer is wrong, he tells the model directly that he does not think it is right and asks it to show the evidence for how it reached that decision. He compares this to calling out a friend or a work colleague, and says asking your AI to prove it should be part of your process, the same as it would be with a doctor.

Q: How does Mark Cuban balance gut instinct against AI-generated data?

Cuban frames trusting your gut as trusting your experience. With large language models such as ChatGPT, Perplexity, and Gemini, you get to feed the system the information that reflects what you already know and understand, so your judgment becomes part of the input rather than something separate from it. When the answer does not look like what it should be, he pushes back and asks for the reasoning and evidence. Intuition, in this framing, is the signal that something is off, and demanding proof is how you resolve it.

Q: What makes a business idea worth pursuing in Mark Cuban's view?

People liking it. Cuban says everyone has millions of ideas and 99.99% of them are junk, so validation matters more than the idea itself. His process starts with telling friends and asking what they think. If they like it, the next step used to be searching Google to see whether anyone had done it before. Now you can ask ChatGPT or Gemini whether anybody has tried the idea and to show who, what, and how it works. If the tools cannot find anyone, it may be a unique idea, and then you ask the AI to write a business plan and keep iterating.

Q: Why does Mark Cuban treat doing things the usual way as a red flag?

Because copying leaves no room to change anything. His quote is that when you see a thousand people doing the same thing, there is no reason to want to be the thousand and first. He says the world does not need another Uber or another DoorDash. The people who change things are the ones who come up with a new idea and have the guts to turn it into a business. He also cites Steve Jobs's line that everything is a remix, meaning you keep adding new ideas to what you have already learned until it leads to a business concept or a change in your life.

Summary & Key Takeaways

• Cuban believes AI will create the world's first trillionaire because the industry and technology are changing so rapidly that someone will find a use nobody has thought of yet. He compares the moment to the early days of PCs, networks, the internet, mobile phones, and 5G, but says AI dwarfs all of them, and the winner could be one person in a basement.

• On a day-to-day level Cuban uses AI for almost everything. He downloaded Replit and typed a plain description of an app to compare pharmacy costs between his company and others with price-change alerts, and it produced a first pass within minutes. He also uses text-to-video tools for the Mavs and other businesses, generating ten seconds of footage from a sentence in about a minute.

• Cuban warns that artificial intelligence is not automatically right, now or ever. He tells ChatGPT when he thinks it is wrong and asks it to show the evidence behind its decision, the same way he would challenge a friend, a colleague, or a doctor. He also flags the pattern of doing something the way it has always been done as a personal red flag.