In January of 49 B.C., Julius Caesar stood on the north bank of a small river in northern Italy. Behind him stood an army. Ahead of him stood Rome, and a decision that would alter the course of history.

Caesar had a choice. If he crossed the river with his forces, he’d be violating the Roman Senate’s order to disband. Once he stepped foot into the water, there would be no path left for negotiation. But if he chose to stay, he’d be forfeiting power that he’d spent a decade consolidating.

Caesar famously crossed the Rubicon, uttering the words alea iacta est. The die is cast. There’s no going back.

Granted, most of us won’t encounter a choice with such massive historical weight. But it doesn’t always feel that way. Everyone faces major decisions over the course of their lives: do you stay at a job, or strike out on your own? Do you save a relationship, or end it? Do you stick with the plan you've built your identity around, or admit it's time for something else? You may not have to physically cross a river, but these decisions have gravity.

In the 20 years since I launched my businesses, I’ve faced many “cross the Rubicon” moments. What I’ve learned is that the hardest part is getting clear about what you actually want before taking the step. It’s this process—separating noise from genuine desire—where AI can help. Here are three ChatGPT prompts I’ve been using to help determine what comes next.

Being a leader means making tough choices, which also means the specter of possible regret is never far away. It’s tempting to think that by staying in one place, you may never have to live with a decision that doesn’t ultimately pan out.

In fact, research has found the opposite . It turns out that people feel stronger regret over action compared to inaction in the short term, but that pattern reverses over time, with inaction becoming the more painful regret in the long run.

I see this pattern all the time. Founders shirk from a good idea because it feels too risky in the moment, then watch a competitor build it two years later. Leaders keep someone on the team past the point they should've made a change, because firing them felt riskier than letting them stay on.

While there’s no failsafe against regret, there are lots of strategies for inhabiting the mind of your future self, like the 10-10-10 rule (how you’ll feel about a decision in 10 minutes, 10 months and 10 years). But lately, I’ve also found that AI makes a helpful thought partner in trying to parse possible options. The next time you’re wondering what to do, try this prompt:

"I'm deciding between [option A] and [option B]. Walk me through how I'd feel in five years if I chose the one that didn't work out. Then do the same for the other option. Don't tell me what to pick. Just map out both possible scenarios and possible regrets."

A lot of times, people think they're stuck on a big decision, when what they're actually stuck on is an unexamined bottleneck underneath it.

Take AirBnB: Back in 2009, the then-fledgling platform was in acute danger of failing. Revenue had flatlined at $200 per week, and its founders had about maxed out their credit cards. Initially, the team thought any solution would have to be scalable—after all, explains co-founder Joe Gebbia, “we believed this was the dogma of how you're supposed to solve problems in Silicon Valley.”

But after spending some time on the platform, they realized the actual constraint wasn’t something that could be solved with code. The problem was that the listing photos were poor quality, and potential renters got no real sense of the homes they’d be staying in. The solution the team came up with was not scalable, but it was simple: rent professional camera equipment and work with listers to upgrade their photos. Almost immediately, AirbBnB’s revenue doubled.

AirBnB got lucky—they located their hidden constraint before they went bust. Generative AI can help save your time and sanity when it comes to diagnosing a hidden problem you may never find. Here’s a prompt to get you started:

"Here's the decision I'm facing: [describe it]. Ask me five questions to figure out what's actually limiting my choice: money, time, fear, someone else's expectations or something else. Don't suggest solutions yet. Just help me find the real constraint."

I’m a big believer in the power of mentors. During the most difficult days of building my startup, I leaned heavily on the advice of others—those who had been there before, and who had the clarity and perspective that I lacked. To this day, I've never made a major decision without running it by someone who'd seen something similar play out before.

While I’ll always recommend seeking the input of people you trust, AI is an incredibly helpful substitute when you’re desperate for advice at 11 p.m. on a Tuesday. Here’s a prompt to get you out of your own head: "Pretend you're a mentor who's seen hundreds of people face this exact decision: [describe it]. What do most people get wrong when they're in this position? What do the ones who end up satisfied with their choice tend to do differently?"

None of these prompts will lead you to the right answer, or tell you what to do. But they can help you to more clearly see the options in front of you. That way, if and when you decide to cross the Rubicon, you can take the step with confidence.