Executive Summary
Fernando Nikolic appeared on the Peter McCormack podcast in June 2026 to share his views on AI and the implications for small businesses. The punchline is that solo-preneurship is more feasible than ever, but the major risk is distraction. Great episode for any aspiring entrepreneur.
Nikolić describes operating a company as its only human employee, using AI agents and tools to build software, process data and perform work that would previously have required a funded team. The discussion covers one-person companies, collapsing software-development costs, and risks.
Most actionable takeaways
AI lowers the minimum viable size for any company. This isn’t a new idea but good to hear this described with an actual example. This does NOT mean you don’t need employees or capital, but can grow to a greater size without needing these inputs than was true previously.
Software is no longer a moat. Cuts two ways: it’s not a moat for software companies, but it’s also no longer a barrier to entry for entrepreneurs. Software is still valuable, but the value is around specific data, specific workflows
Abundance creates a focus problem. From personal experience, it’s now easy to spin up 10-20 vibe coded apps that are ~80% functional. But very hard to decide which ONE to go far with.
Actionable insights:
Focus. AI enables you to start - but perhaps not finish - an infinite number of projects. This makes you feel productive while getting nowhere. So pick ONE. One idea, one workflow. Nail it and scale it.
Build around a scarce asset—not merely AI capability
Do not position the company as “we use AI.” Almost everyone will use AI. Start with something competitors cannot immediately reproduce:
Exclusive or difficult-to-assemble data
A trusted niche audience
Embedded customer workflows
Regulatory or operational expertise
A proprietary sourcing or distribution channel
Use AI to delay hiring, not to eliminate human judgment
A founder should first deploy AI across research, coding, customer support, marketing operations and internal analysis. Hire only after identifying work where human relationships, accountability or specialized expertise clearly outperform automation.
This can significantly reduce the capital required to test a business.
Pick one compounding workflow.
The strongest one-person businesses will probably not be collections of disconnected experiments. They will have a reinforcing loop:❝Create useful output → attract users → collect proprietary feedback or data → improve the product → create better output.
For a media or research business, the moat may not be the individual summaries. It may be the accumulated, structured archive and the judgment applied across hundreds of sources. For service businesses, this may be delivering time to value or best customer experience. Solve for ONE thing that matters in an end market. Grow and scale from there.
The Big Idea
Solopreneurship is possible, but focus is required as AI perversely makes distraction more likely. Pick one business, one idea, and one workflow to nail and scale.
“The Peter McCormack Show” is the podcast, from June 2026. Find the full episode:
https://youtu.be/pOrXjz-lLV0?si=VnHpqoDA9zoCV4xj
Which topics should we write more about?
Was this issue useful?
Until next time,
Compounder
