Henrik Werdelin, co-founder of NYSE-listed BARK and venture studio pioneer at Prehype, reveals why AI alone can't build successful businesses—and introduces "Donkeycorns," profitable million-dollar companies run by just one or two people. He shares groundbreaking insights on relationship capital, the critical human elements AI can't replicate, and how entrepreneurs can leverage AI agents while maintaining the authenticity and authority customers crave. Discover practical AI workflows, from running multiple autonomous agents simultaneously to using voice-to-text for thought partnership, plus why "thin data" threatens to commoditize everything AI touches.
Co-Founder & Chief Strategy Officer | BARK (NYSE: BARK)
Henrik co-founded BARK (formerly BarkBox) and led the company from a subscription box concept to a publicly-traded NYSE-listed company serving over 2.5 million dogs monthly, bringing joy to dogs and their people through innovative products and services. He currently oversees strategy, brand development, and M&A while maintaining the company's mission to be "Disney for dogs," demonstrating how customer-founder relationships drive business success beyond simple utility.
Founding Partner | Prehype
Henrik co-founded Prehype, a pioneering venture development firm and startup studio headquartered in New York with offices in London and Copenhagen, that co-creates new ventures with leading VCs and corporations. Through Prehype, he has helped launch and scale multiple successful companies including Ro (digital health), Managed by Q (acquired by WeWork), AndCo (acquired by Fiverr), and numerous other ventures, working with major corporations like Coca-Cola, LEGO, Microsoft, Danone, GE, and Mondelez.
Co-Founder | Audos
Henrik co-founded Audos, an AI-powered platform that democratizes entrepreneurship by helping everyday people build million-dollar businesses without technical skills, raising $11.5 million in seed funding led by True Ventures. Audos has already helped launch hundreds of "Donkeycorns" (profitable one- to two-person companies generating $1M+ in revenue) and aims to scale to launching 100,000+ companies annually, fundamentally changing how small businesses are created in the AI era.
Donkeycorns Defined | Henrik reveals his concept for million-dollar companies run by two people or less, celebrating profitable solo entrepreneurship over unicorn obsession.
Relationship Capital | When AI can build everything, the competitive moat becomes depth, density, and durability of customer relationships that only humans can create.
AI Founder Experiment | Henrik's five-year journey attempting to make AI the founder itself, discovering why authenticity and authority require a human in the loop.
Thick vs Thin Data | The critical difference between superficial data AI processes easily and the nuanced human understanding that creates real connection.
Customer-First Entrepreneurship | Why identifying who you want to serve for ten years matters infinitely more than having a clever business idea.
Multi-Agent Workflows | How Henrik runs five to six AI agents simultaneously on his home computer while managing them remotely from his phone.
Context Engineering | Using text files from books and personal knowledge to guide AI beyond average responses toward uniquely valuable outputs.
Prompting as Architecture | Designing AI interactions like Steve Jobs designed hallways, creating specific paths that generate unexpected but directional results.
Mental Diabetes Warning | The danger of accepting AI's first good-enough answer instead of pushing for truly excellent creative work.
Voice AI Integration | Leveraging Grok's less agreeable voice mode during drive time as a thought partner for productive ideation.
00:00 - 04:35 | Introducing Donkeycorns: The Anti-Unicorn Business Model
04:36 - 09:20 | How AI Democratizes Entrepreneurship for Everyone
09:21 - 14:15 | Why Humans Must Stay in the Loop
14:16 - 19:40 | Depth, Density, Durability: The Relationship Capital Framework
19:41 - 24:55 | Thick Data vs Thin Data: What AI Can't Understand
24:56 - 30:10 | Choosing Your Customer Over Your Idea
30:11 - 35:25 | Running Multiple AI Agents Simultaneously
35:26 - 40:40 | Advanced Prompting Techniques and Context Engineering
40:41 - 45:43 | Voice AI, Health Tracking, and Productivity Tools
"Donkey corn is a company that does a million dollar in turnover with two people or less, and so it's the anti unicorn, but it still grinds like a mule, but it parties like a unicorn."
"We got convinced that you needed a human in the loop, you needed somebody to facilitate the link between the customer and the founder, and that has to be a human for basically two reasons: one is to create authenticity, and the other one is to have authority to basically say, hey, I'd like to solve that problem for you."
"Relationship capital is going to be one of the last moats that's going to be available when everybody can do everything."
"Thick data is the data that you as a human compute that makes you understand when I say the party is just getting started or the mood in the office is just kind of off, like these are all these subtle things that we pick up that we as humans have been trained through tens of thousands of years to basically compute."
"If everybody is obsessed about building unicorns and technology is being democratized, maybe there's a new asset class of entrepreneurship that doesn't get a lot of attention."
"The more that I learn about AI, the more I learn that this is all about understanding humans, and I think one of the interesting questions with these things is not how intelligent or how amazing is AI, it is really how amazing and how unique are humans."
"I'm increasingly thinking about mental diabetes, like how do I not just go for good enough, because I have a lot of stuff going on, and sometimes when AI gives me an option, I'm like, yeah, that's pretty good, but really I'm worried that I kind of do the same thing that social media did to me where you're just doom scrolling, so you're doom ideating."
"True innovation, true exponential creative stuff comes from systems of open endedness, by pursuing with great intensity interestingness, and that creates stepping stones, and you don't know where those stepping stones will take you, but if you're lucky, it's going to end up with something great."
"Narratives are like source code in the LM world, you need stories because the stories are what the LMs really understand very well, and so the more that you can make something that is pretty unique, the more you get a unique output."
"There is this very unique period of time where people that understand AI and can use this, they just have a very big head start against everybody else, because there are people that know how to put on this Ironman suit and there are people that do not, and the people that do just seem to be able to do ten people's work."
Accelerate your productivity & creativity using 47 custom prompts to grow your income, success & quality of life by downloading our free “AI Accelerator Pack” HERE.
START HERE
