Philosophy Eats AI
"If software is eating the world and AI is eating software, what eats AI?” Michael Schrage and David Kiron pose this question in their MIT Sloan piece, Philosophy Eats AI. Their answer - philosophy - may sound academic. It isn’t. It’s the key to extracting real business value from artificial intelligence.
Most organisations treat AI as a technical upgrade. They build models, integrate APIs, experiment with prompts - and hope for the best. But behind every model is a question few executives stop to ask: What is it learning?
If we want AI to do more than generate plausible sentences or summarise spreadsheets, we must give it something deeper: a philosophy of the enterprise. That philosophy isn’t abstract. It’s practical. It lives in your data, your metrics, your labels, your architecture. And if you don’t define it, your competitors - or worse, general-purpose AI vendors - will.
🔵 Vocabulary of Value Creation
Every business has a unique way it creates value. For Starbucks, it’s loyalty. For Amazon, it’s optimisation. For your business, it might be trust, efficiency, creativity - or all three.
But here’s the problem: most organisations have no structured way to talk about this. They lack what Schrage and Kiron call the vocabulary of value creation - the set of core concepts that define how and why the business works. This isn’t about mission statements. It’s about semantics. 'hardcore semantics' that have been formalised into an ontology - a machine-readable structure that AI systems can reason over.
This is not optional. Generative models don’t simulate intelligence by replicating logic. They do it by recognising and encoding latent patterns. If your business hasn’t surfaced its core logical structure - if your semantics are buried in slide decks and siloed spreadsheets - your models are learning on noise.
🔵 Ontological Core
This leads to what I call the ontological core: the set of concepts that sit at the nucleus of your enterprise’s identity. These concepts are discovered through use cases that generate real value and by competency questions that test their relevance.
Once modelled, this core becomes a lens for focused AI reasoning. It’s the part of your business that you want your models to understand deeply and protect fiercely.
🔵 Semantic Data Products
These concepts MUST connect to real data. Enter the semantic data product. Based on the open DPROD specification, this approach treats data not as raw material, but as a product: something with value, cost, stewardship, and clear ownership.
In short, DPROD connects your philosophy to your data architecture. It creates a distributed, AI-ready knowledge graph where every dataset knows what it is, why it matters, and how it fits into the broader picture.
The real opportunity isn’t just in cleaning data or building models. It’s in initiating a virtuous cycle: You use AI to help define your ontology, and your ontology defines how your AI thinks and the value that it creates.
