Billionaire Mark Cuban Said He’d Trust a Seeing-Eye Dog Over a Self-Driving Car, ‘A Dog Will Recognize Adversarial Situations Better Than AI Can’

Barchart
Abrir en Barchart
Billionaire Mark Cuban Said He’d Trust a Seeing-Eye Dog Over a Self-Driving Car, ‘A Dog Will Recognize Adversarial Situations Better Than AI Can’

Mark Cuban is one of the loudest voices telling businesses that ignoring artificial intelligence (AI) is a fatal mistake. But there is one job he would not hand to it: Walking a blind person down the street.

In a WIRED interview in 2024, Cuban had a choice: a self-driving car or a seeing-eye dog. Cuban picked the dog. “I would trust the dog,” he said. “Because a dog can sense issues. Nothing about a self-driving car understands what’s adversarial or not.” In the same conversation, Cuban put it more bluntly, saying that he thought smart puppies were smarter than AI at the time or in the near future.

Can’t Get Enough Options?: Join the list for Barchart’s daily unusual options report, delivered free.

 

That is a striking thing for him to say, because Cuban's public position on AI runs almost entirely the other way. Barchart has previously reported on Mark Cuban warning that businesses ignoring AI are already falling behind and that people ignoring it are like internet-deniers, about as dismissive a label as he has for anyone. Cuban has also walked through several ways the AI industry could collapse while remaining bullish on it, and more recently named the software he thinks will be the next great AI application.

Sponsored Content: A $9 Trillion Flying Car Market Is Taking Shape. This Company Is Open to Investors.

Cuban's bullishness is about AI as a business tool — doing drafting, summarizing, pattern-finding work with a wide margin for error and a human reviewing the output. His skepticism is about AI in an adversarial physical environment, where the margin for error is zero, and the situation is actively trying to surprise you. A guide dog does not have better pattern recognition than a modern perception stack, per se. Rather, it has better judgment about the unfamiliar and can adapt more quickly as the environment changes. 

Importantly, “adversarial” is a technical term here, so it doesn't necessarily mean hostile. In machine learning, it describes inputs specifically shaped to break a model. The traffic sign with tape on it that a classifier reads as a speed limit change, the edge case that never appeared in training. It remains an unsolved problem. In that realm of thought, Cuban's comment makes a bit more sense than saying puppies are smarter than AI. He is simply saying that most living things can adapt better to the unknown because AI can typically only react to things it has been taught or seen before.

Contrast that with billionaire Michael Saylor, who argued the opposite end of the same question — that self-driving cars will be obviously safer than human drivers, that an unintelligent appliance will soon seem absurd, and that a 25-year-old should skip an expensive degree in favor of a $20-a-month AI subscription. Two billionaires, both bullish on AI as an economic force, drawing the reliability line in completely different places.

The industry has kept moving in the interim. Alphabet (GOOGL) has continued expanding Waymo's commercial footprint, and Tesla (TSLA) has kept pushing its own autonomy timeline. Cuban's remark is a data point about how a prominent investor thought about machine reliability, not a verdict on where the technology now stands.

One caveat matters more than the quote. The WIRED interview is roughly two years old, and Cuban's view may have moved, as he has made several fresh AI comments since. Anyone reading this as his current position on autonomous vehicles is reading it wrong.

What has not aged is the framing. The question Cuban was answering was not whether AI is impressive. It was whether he would bet someone else's safety on it, with no human in the loop, in a situation designed to go wrong. He said he would take the dog.


On the date of publication, Caleb Naysmith did not have (either directly or indirectly) positions in any of the securities mentioned in this article. All information and data in this article is solely for informational purposes. For more information please view the Barchart Disclosure Policy here.

 

More news from Barchart

Billionaire Mark Cuban Said He’d Trust a Seeing-Eye Dog Over a Self-Driving Car, ‘A Dog Will Recognize Adversarial Situations Better Than AI Can’ Institutional Investors Invest in Long-Dated Taiwan Semiconductor Call Options - A Bullish Signal on TSM Yum China Stock Is Statistically Terrible. Here’s What Wall Street Isn’t Telling You. Costco Earnings Bull Put Spread has a High Probability of Success