For OpenAI CEO Sam Altman, dropping out of college to launch his first startup, Loopt, wasn’t as big a risk as it might seem.

In fact, in Altman’s telling, the decision to leave Stanford in 2005 was not particularly calculated, but rather somewhat unexpected.

“This was not like my life plan, but this other opportunity came up,” Altman said during an interview at his private grade school John Burroughs School in St. Louis. “It seemed like a really fun thing to try, and importantly I realized that I could go back. I think that’s the key to most risk. Most things are not a one-way door. You can try something, [and] if it doesn’t work out, you can undo it.”

In this case, Altman could have always gone back to college and earned a bachelor’s degree if things with Loopt didn’t pan out. Altman was just 19 years old at the time when he left Stanford to start Loopt, which offered an early version of location services. The startup turned out to be a moderate success: The technology proved more than promising, with location services soon becoming a critical component of virtually every mobile application from banks to games to news, but Loopt never found its footing. The company eventually sold to the banking firm Green Dot in 2012 for $43.4 million, which notably put Altman on the map in Silicon Valley.

So the risk for Altman ultimately paid off. Though he doesn’t necessarily see the decision to launch a startup as all that risky to begin with, society at large is often risk-averse, making his decision seem more drastic than it should have been, according Altman. “Generally I think people are horribly miscalibrated on risks,” he said.

Leaving college to pursue a startup wasn’t wasn’t such a big risk in his mind because the traditional career path of the 1950s and 60s—in which someone went to college, got a job, and stayed at the same company for the majority of their career—was no longer as rewarding as it was in the past, according to Altman.

“Then it gradually stopped working,” Altman said. “Now I think the traditional path is, I won’t say falling apart, but it’s quite challenged. And AI will probably disrupt things even more and put even more variance in the traditional path.”

Current-day career paths offer more opportunity for self-correcting mistakes, Altman added. There is substantial evidence that millennials and Gen Z workers are much more open to switching jobs than previous generations and don’t attach any negative stigma to doing so. During the booming job market of the pandemic, switching jobs became extremely lucrative. But even the current cooling of the job market has done little to temper the restlessness of younger employees and their knack for eyeing other jobs.

With that in mind, Altman encouraged the audience of high school students to consider risk as opting for too much stability. “In what is now a very dynamic world, the risky thing is to not go try the things that might really work out,” he said.

Altman also warned against the regret that can come with an overly cautious approach. Without any risk, “you kind of look back at your career, 10, 20, 30 years later and say, ‘Man I wish I had tried the thing I really wanted to try,’” Altman said. “You should put a huge premium on doing that anytime you feel like you might say that later.”

Mindful of his audience, Altman did make one thing clear. “Please don’t go home and tell your parents I’ve told you to drop out of college,” he joked.