Yes, Even in the Age of AI You Still Need a Co-Founder
Entrepreneur Office Hours - Issue #314
Inside the Office
Last week, we hosted something called Startup Fest. It’s our annual Spring semester networking event where hundreds of students across Duke’s entrepreneurial ecosystem show up with ideas, half-ideas, résumés, curiosity, and a healthy dose of social anxiety. It’s loud. It’s awkward. It’s energizing. And it’s one of my favorite reminders of how entrepreneurship actually begins. Despite everyone’s obsessions with products and pitch decks, entrepreneurship starts with people trying to find each other. And that’s exactly what Startup Fest is for.
Toward the end of the event, I bumped into a student who’s in the early stages of launching a startup. He was sharp, thoughtful, and clearly frustrated. He’d been struggling to find a cofounder and was starting to convince himself it wasn’t necessary.
“Honestly,” he said, “with all these AI tools now… do I even really need a cofounder?”
I didn’t hesitate. “Yes. You absolutely do.”
That answer usually surprises people, especially right now. AI has made doing things by yourself easier than ever. You can design, code, research, write, analyze, and automate at a level that would’ve required a small team not that long ago. In that context, a cofounder might seem obsolete, but none of those things are the main job of a cofounder.
A good cofounder isn’t just someone who helps you execute. A good cofounder is someone who helps you focus by bringing complementary strengths. If you’re great at vision and customer conversations, you need someone who thrives in systems and execution. If you love building, you need someone who loves selling. Not because AI can’t help, but because humans still make judgment calls, prioritize tradeoffs, and decide what actually matters.
But even that’s not the most important reason.
Startups are hard in ways that don’t show up on product roadmaps. They’re lonely. They’re emotionally exhausting. They involve long stretches where nothing seems to work and no one is cheering for you. A cofounder is someone who goes to war with you… someone who can say, “This week sucked, but we’re not quitting,” and actually mean it. AI can optimize workflows. It cannot sit across from you and keep you moving when motivation collapses.
Plus — and this is a harder truth founders don’t always like hearing — finding a cofounder is a test. If you can’t convince one other person to commit years of their life to building this thing with you, how the heck do you expect to convince customers? Investors? Early employees?
To be clear, I don’t mean solo founders can’t succeed. Some do. But the difficulty of finding a cofounder isn’t a bug in the system. It’s a feature that forces you to articulate your vision clearly, listen, and… yes… sometimes even compromise. And it does all that while forcing you to prove — early — that your idea is compelling enough to inspire belief.
So yes, finding a cofounder is hard. But remember, it’s supposed to be. If finding a cofounder were easy, it wouldn’t be one of the most meaningful filters entrepreneurship has to prove you’re building something the world needs.
-Aaron
Worth Your Time
Founders love the lightning-bolt story: one insight, one pivot, one launch that flips the switch. This Fuqua Insights podcast episode — “Can 1% Improvements Transform Your Business?” — is a calm (slightly ruthless) counterpoint: the startups that win more often build a repeatable way to get a little better — over and over — on purpose. In other words, rather than trying to be brilliant, focus, instead, on getting a little better each day.
In the podcast, Duke Fuqua professor Sharique Hasan looks at how experimentation (often via A/B testing) becomes a pathway to compounding improvement. And it turns out to be especially important when you do it consistently.
By the way, one of the biggest value adds of Duke I&E — and why we're excited to keep expanding this newsletter — is that we’re embedded in a research university, which means the “how entrepreneurship actually works” lessons don’t just come from war stories. They also come from colleagues who study these questions rigorously and at scale, and we get to use spaces like this to translate that kind of evidence into practice.
Tools We’re Tinkering With
-Aaron Dinin
This week’s tool is simple, free, and wildly underused: Vidyard. Specifically, the Vidyard Chrome extension that lets you record a quick video of yourself and drop it straight into an email.
Most entrepreneurs default to walls of text when they need to explain something — a follow-up, a clarification, a pitch, a thank-you. But text strips out tone, context, and humanity. A 60-second video does the opposite. It lets people see your face, hear your voice, and understand intent immediately. In practice, that means fewer misunderstandings, faster alignment, and better relationships.
I used to love using Vidyard back when I was selling because it was a frictionless way to keep my face in front of potential customers. Now I actually use it to email students because they’re much more likely to actually watch video instructions than read anything I write in an email.
And by the way, Vidyard isn’t the only option. We’re not trying to sell you anything here… just being helpful. The important thing is sending videos rather than long emails. And Vidyard just happens to make that really easy. You click record, talk like a human, and paste the link. Simple.
By the way, I realize the irony of telling people they should be sending videos instead of long, text-heavy emails in a text-based newsletter. But even if some audiences prefer emails, most people default to video whether we like it or not, and the founders who win are the ones who communicate faster, clearer, and more humanly.