Brian Kerns can’t pinpoint the exact moment pitch decks began to look the same. Somewhere in the last two or three years, the shift arrived: more polish, more sophisticated graphics, tighter branding, and less variation from one deck to the next.
“Now I assume all deal room documents have likely been engineered by AI,” Kerns says, joking. “There is absolutely a point where I feel like we’re all converging on the exact same optimized pitch deck.”
For Kerns, the feeling calls to mind a line from Rush’s 1981 song Red Barchetta: “Straining the limits of machine and man.” It’s a fitting frame for a question he keeps coming back to: As AI makes it easier to optimize every element of a pitch, what gets lost in the process?
Kerns, a startup investor and advisor with Unventure Capital and SWAN Impact Network, serves as a mentor-in-residence for VentureWell’s Aspire Climatetech and Ocean Enterprise Accelerator programs, where he coaches early-stage founders through the hurdles that can determine whether a promising new venture catches the interest of a larger investor, aligns with a strategic partner, or optimizes for a profit-based growth model.
Break the Mold
Kerns sees hundreds of pitch decks each year across a range of industries, and lately more of them seem overengineered. They have eye-catching graphics and a tightly coordinated color palette, but they’re light on the specifics that make a founder’s story their own. The polish, he’s found, often masks an absence.
He is careful to note that artificial intelligence (AI) didn’t start this. He says that pitch decks were already gradually drifting toward a sea of sameness before Claude and ChatGPT ever entered the picture. For years, accelerator programs stressed the importance of building a professional deck, most of them prescribing a similar format. For a while, that system worked: A polished deck signaled thoughtful effort, proved that founders had internalized their research and stress-tested their assumptions, and verified they had prepared for the presentation.
“AI didn’t invent this uniformity,” Kerns says. “It just made it free and instant.” But once every founder had access to the same easy polish, the signal lost its meaning. A deck can look right while lacking depth and substance. “Often the decks seem to lack a soul,” Kerns says. “They’re lifeless.”
He’s describing what researchers have dubbed “the sameness trap”: the homogenized output of overreliance on AI. The result is diluted distinctiveness, one pitch deck indistinguishable from the last. Forgettable. “It sounds funny, but it may be too clean,” says Kerns. “Too optimized.”
Even the strongest decks can trigger a negative reaction in Kerns now: “Wow, somebody or some system did a great job engineering this for you.” His advice to founders now runs against pitch-deck orthodoxy. “I’ve actually started telling companies I advise: Break the mold,” he says. “Instead of using the default industry template, don’t. Do something different. Convince me that you put thought into this and it reflects your personality and passion. Change the order.” The old rule—hit the right 10 to 12 slides in the exact right order—no longer holds.
Prioritize Authenticity
For Kerns, the differentiator is soul and authenticity. “Really good founders tell you a story,” he says. “And often it’s a personal story.” The strongest pitches, Kerns argues, bring an investor inside that story—not just the market opportunity, but why this founder, specifically, is the right person to pursue it. Which is precisely the part AI tends to omit or flatten.
“AI is not going to reach into your soul,” Kerns says. “Not yet.”
Even the most thoughtful prompts can’t produce the details that make one founder’s story impossible to mistake for another’s, as those come only by hard-earned, lived experience. Kerns’ advice to break the mold goes beyond slide order. It’s about how much of themselves founders are willing to put into their pitch, and whether they’re willing to let a chatbot sand off the parts that make them memorable.
Don’t Outsource Your Thinking
But Kerns is no AI skeptic. He’s clear-eyed and pragmatic. Optimistic, even. And he’s blunt that using AI isn’t optional anymore, so long as a founder doesn’t let it think for them. “Frankly, if you’re not using AI, you’re not competitive,” he says. “The problem is, you’re also not competitive if you let it take over and you quit thinking. Never outsource your ability to think critically.”
What he wants isn’t less thinking, but better thinking, done faster. “I want people who can think critically and use AI to drive progress,” Kerns says. “Not people who outsource their critical thinking. That is a 100% fail.” The line he draws is narrow: AI should make founders faster and better informed, not replace their judgment. “Any place you use it to make you more efficient and get better information—everyone has to buy into that philosophy,” Kerns says. “You simply have no choice.”
His advice to founders trying to move quickly without losing authenticity is to go slow on purpose. “Ease into it. Don’t just start turning things over to AI,” he says. “Understand where you have to be authentic when you present and when AI can provide the data-driven backing you require.”
And, above all, don’t let AI stand in for the parts that need another human in the room. “Do not let it replace the engagement for you, because at some point, due diligence is coming for you,” he cautions. “And we’re going to sit down and have a conversation. We’re going to ask a bunch of questions you may not have thought to ask AI to answer.” That conversation, Kerns says, is where the gaps show up—or don’t. A founder who has done the work reveals it under questioning; a founder who has outsourced the thinking reveals that too.
Be Ready To Answer For It
“You have to be honest. You have to be willing to back everything up,” Kerns says. “You’re accountable. AI isn’t.”
He expects that accountability to become more formal soon, predicting that investors will start requiring founders to disclose where they used AI. “In one of my startups, the thought leaders are proposing that you have to disclose AI use and show your prompts,” Kerns shares. “I’m seeing people at all levels stop responding with their own personal words and instead take the shortcut of pasting output from Claude rather than writing their own responses.” Kerns advocates for disclosure as a form of accountability: If you can’t show your work, you probably didn’t do it. If you didn’t do it, you probably can’t defend or explain it.
Keep Doing the Work
What all of this is really protecting, in Kerns’ telling, is a kind of diligence AI can’t shortcut: knowing your market well enough to survive unscripted conversations and market deviations.
“Customer discovery is not about asking the first question,” he says. “It’s about the fifth question that comes out of the discussion that you never knew to ask up front.” That’s the human part of the process, he argues, that no amount of research or technology can replace. A founder might think they’ve found the right person to talk to, only to learn three conversations in that the person who actually understands how to move the sale forward sits in another seat.
AI can’t substitute for finding that person, and when a founder skips that step, Kerns says it shows. He sees plenty of AI-generated output that is decidedly wrong—confident and well formatted, perhaps, but built on a shallow, lazy read of the market. “In those cases, it’s garbage in, garbage out,” Kerns says.
Investors, in the end, are underwriting a founder’s understanding of people and markets, which AI can help sharpen but never originate. “There’s no black and white on this,” he says. “But anytime you’re misrepresenting something, or using it to short-circuit a process where you really need to be engaged, that’s where I get uncomfortable.”
The New Baseline
Kerns gets animated talking about cost and speed. He doesn’t mince words about what he now expects from a software founder: “If I have a founder in the software space, and their team isn’t using AI to accelerate their time to code, time to market, time to test, that’s a fail. Right there. Done.”
And it’s not just software. He points to hardtech founders using AI to accelerate 3D-printed prototyping, optimize sensors, and streamline supply chains. He pushes back strongly on the dated assumption that hardware development and traditional operations can’t move at the speed of agile software.
The founders who stand out to him, across all segments, are the ones using AI to compress what used to be slow and expensive: a software team of two or three, doing what once took the work of eight or 10, or three or four hardware iterations tested against the market in the time it once took to develop one simple version.
He’s watching a newer pattern too: founders using AI to build their own back-office systems, marketing infrastructure, sales operations, the stuff that used to mean six figures spent on outside vendors before a product ever shipped.
Cut the Timeline, Not the Corners
It’s still playing out in real time, but Kerns says the founders doing it well aren’t cutting corners. “They’re not bypassing their responsibility to produce good information, engage the market,” he says. “But they’re moving blazingly faster and much leaner.”
AI has changed the baseline for what founders can do quickly and cheaply. But when everyone has access to the same tools, the tools stop being the differentiator or the defensive “moat,” and the advantage moves elsewhere. What separates one founder from another is what they do with the time they save: the conversations they have, the questions they pursue, the customer feedback they uncover, and the experience they bring to the story.
“I still believe there’s no replacement for going out and doing the work yourself,” says Kerns.
1. Break the mold. Don’t default to standard pitch-deck templates. Make due diligence materials your own.
2. Prioritize authenticity. AI can polish your presentation, but it can’t supply the experiences and perspective.
3. Don’t outsource your thinking. Use AI to work more efficiently, but keep the thinking yours.
4. Be ready to answer for it. You’re accountable for whatever AI helps you produce.
5. Keep doing the work. Talk to customers. Ask the third, fourth, fifth questions. Follow unexpected leads.
6. Cut the timeline, not the corners. Let AI cut time and costs, not your responsibility.