What if everything you've been told about AI implementation is backwards?

In the latest episode of the OmnichannelX podcast, host Noz Urbina sits down with Lasse Rindom, and they shatter the myth that AI is your secret weapon. Spoiler alert: it's not. With 127 million people using ChatGPT daily, AI has already become as common as Excel. The real question isn't whether you should use AI, but how to stop treating it like a magic wand and start building actual business value.

Through 67 episodes of interviewing AI leaders, Lasse has discovered a pattern: companies are failing because they're asking "what can AI do?" instead of "what do we want to achieve?" From exposing why your million-dollar AI investment might be worthless to revealing how "brownfield thinking" can save your transformation, this conversation flips conventional wisdom on its head. You'll discover why context engineering beats prompt engineering, how every business process is secretly about metadata, and why the Wright Brothers' invention of the airplane tells us everything we need to know about where AI is headed.

Whether you're a CEO wondering why your AI initiative isn't delivering ROI or a practitioner trying to move beyond chatbot experiments, this episode delivers the tough love and practical wisdom you need to succeed in 2025's AI reality.

"If you don't start with outcome and outcome discussion about what you want, then you're not gonna end up with outcome. And that has nothing to do with AI at all." – Lasse Rindom

Key Findings

• AI is a commodity, not a differentiator - With 127 million daily ChatGPT users, the technology itself won't provide a competitive advantage; innovation on top of AI will
• Process definition precedes successful AI implementation - Software vendors must own and define processes clearly for AI agents to navigate effectively within digital infrastructure
• Brownfield reality trumps greenfield fantasies - Organisations must work with existing systems and constraints rather than imagining clean-slate implementations
• Context engineering > Prompt engineering - Pre-prompting and establishing proper context matters more than individual user prompts for reliable AI performance
• Metadata production is the essence of business processes - Every business process essentially adds metadata to transform inputs into outputs, making AI particularly suited for structuring unstructured data
• Human change is the limiting factor - Technology adoption speed is constrained by how quickly humans can adapt their mental models and processes
• Measurement from day one is critical - AI initiatives without clear KPIs and success metrics become expensive experiments rather than business improvement

00:00 Introduction and guest welcome

04:26 AI's role in business and value creation

08:05 "AI doesn't need a push - it's already here"

15:40 The importance of outcome-based AI implementation

18:29 Greenfield vs. brownfield: "The world is brownfield"

23:33 From playground money to real ROI

28:57 Beyond generative: AI as restructuring tool

33:38 "Every process is metadata production"

35:16 Reducing entropy: the true purpose of business

40:34 Context engineering vs prompt engineering

46:56 "The Wright Brothers didn't invent the airline"