Hunter Harris from Hunter Software Consulting and I sat down to shoot the breeze. Like me, Hunter is focused on helping companies navigate growing pains. Too small to be big, too big to be small.
Topics Link to heading
It wasn’t meant to be purely about AI, but inevitably these days we ended up dedicating a good bit of our conversation to it.
Besides AI we also touched on
- A quick detour in VC investment perspectives and math, and how founders should interpret that when they think about funding
- What makes a good founder
- Why founders are sales people, whether they like it or not
Full episode here: https://www.youtube.com/watch?v=xKOTGjSsjiw
AI Summary Link to heading
On AI Strategy (Pragmatic vs. Hype) Link to heading
The “Harness” Concept: Their advice to focus on the testing harness rather than the code generation is excellent. In professional software engineering, the bottleneck is rarely typing the code; it’s ensuring the code is secure, maintainable, and correct. Relying on AI to “vibe code” without a rigorous testing framework is a recipe for technical debt.
Proprietary Data: They rightly point out that “GPT wrappers” (apps that just provide a UI for someone else’s AI) have no “moat.” If OpenAI or Anthropic adds your feature to their next update, your business vanishes. Real value now lies in having data that the big models haven’t seen yet.
On Fundraising (The “VC Math”) Link to heading
Alignment of Interests: Ardinois provides a very honest look at how VCs think. Many founders fail because they don’t realize that a “successful $10M business” is a failure to a VC. Understanding that VCs are looking for fund-returning “home runs” helps founders decide if they should seek VC money at all, or if they should “bootstrap” (self-fund).
Market Size: His point about the “knife fight with Salesforce” is a classic warning. Don’t try to out-feature a giant; find a niche they are too big to care about.
On Leadership (The Founder’s Trap) Link to heading
Decoupling: The advice on “decoupling” the founder from the business is the gold standard for scaling. If a company can’t function while the founder is on vacation, it isn’t a company—it’s a job.
The “Sales” Mandate: Harris is correct that a founder who “doesn’t want to do sales” is at a massive disadvantage. In the early stages, customers buy the founder’s vision, not just the product. You cannot delegate the soul of the company to a hired salesperson in the first year.
On Education & Careers Link to heading
Understanding the “Black Box”: Their suggestion to build a simple neural network from scratch is great advice for anyone in tech. It removes the “magic” and helps you understand the limitations of AI (like why it hallucinates or struggles with math), which makes you a better prompt engineer and strategist.
Summary Verdict Link to heading
This is sophisticated advice. It is particularly useful for:
- Technical Founders who struggle with the “business” side of things.
- Aspiring Entrepreneurs who need to understand that AI is a tool, not a business model.
- Junior Developers looking to understand how the economic landscape is shifting beneath them.