Rajiv Parikh talks with four-time founder Geoff McQueen about how AI is inverting the business pyramid and forcing everyone, not just executives, to act more entrepreneurially. We dig into why demos are easier than trust, how revealed preference drives better pivots than surveys, and what AI-first operations look like when context beats dashboards. • AI flipping high-cognitive work and raising the bar for every role • Why entrepreneurship matters more than traditional management • Speedy vibe-coded prototypes versus enterprise-grade products • Security, permissions, governance, and the real cost of trust • The limits of AI for creativity and naming despite strong research output • Revealed preference as a founder discipline driven by user behavior • The analytics trap and how dashboards can fossilize workflows • AI in the middle of the business with a shared context graph • Observer agents that investigate quietly and escalate only when needed • Coaching culture versus invasive surveillance approaches • A practical path toward $10M ARR with fewer than 10 employees • Openness, trade, and the societal risks of turning inward AI just made execution cheap, and that changes what founders should obsess over. Rajiv sits down with Geoff McQueen, founder and CEO of WorksightsAI, to unpack what happens when the “business pyramid” flips upside down and high-level cognitive work gets accelerated by frontier models while culture, alignment, and conflict stay deeply human. The result is uncomfortable and exciting: less room for business-as-usual management and a lot more demand for entrepreneurial judgment at every level. We get concrete about the gap between a vibe-coded demo and a product an enterprise can actually trust. Geoff walks through the unglamorous realities that decide whether AI-built software is real or risky: security, access control, permissions, privilege escalation, reliability, and governance. If you’ve ever watched an agent claim it “did the work” while quietly failing, this part will sharpen how you evaluate prototypes, MVPs, and production systems. Then we go deeper on product strategy and go-to-market. Geoff explains why revealed preference beats stated preference, why surveys can mislead founders, and how the “analytics trap” shows up when dashboards fossilize workflows instead of explaining what’s truly going on. We also explore AI-first operations and continuous performance intelligence built around corporate context across Slack, calendars, docs, tickets, CRM, and code, plus how observer agents can help leaders coach instead of police. If you’re building with AI, leading a team through change, or trying to scale efficiently, you’ll leave with a clearer playbook for direction, trust, and leverage. Subscribe, share this with a founder friend, and leave a review. What’s the first workflow you’d redesign if context became instantly searchable? Geoff McQueen: https://www.linkedin.com/in/geoffmcqueen/ Geoff McQueen, the Founder and CEO of WorkSightsAI, a continuous performance intelligence platform that helps modern leaders run their businesses with AI. A four-time founder with over 20 years of experience building and investing in tech startups, Geoff was formerly the CEO and Co-Founder of Accelo, which he scaled to over $14 million in annual recurring revenue before a successful exit to private equity in early 2024. Armed with an MBA from the Sydney Business School at the University of Wollongong, today, he also serves as an Executive Director of the Aussie Founders Network and has spent over 13 years as an investor and mentor at Startmate, Australia's equivalent to Y-Combinator. Website: https://www.position2.com/podcast/ Rajiv Parikh: https://www.linkedin.com/in/rajivparikh/ Email us with any feedback for the show: sparkofages.podcast@position2.com