Executive Summary
Most industrial companies are adding AI to their existing models and calling it transformation. Professor Venkat Venkatraman of Boston University’s Questrom School of Business has a name for this: digital paint. In this episode of Arkaro Insights, the co-author of Fusion Strategy explains why the $75 trillion asset-heavy world, chemicals, food, agriculture, industrials, is the real frontier of the intelligence age, and why the companies that will win are those that design intelligence into the core of what they do rather than bolt it onto the edges.
Core insights:
- The age of intelligence has three foundations: digital AI (language, images, sound), physical AI (robots, autonomous vehicles, drones), and industrial AI (self-optimising factories and supply chains) — all orchestrated by human intelligence, which remains irreplaceable
- Roughly $75 trillion of global GDP has yet to be systematically digitised — this is the asset-heavy world of chemicals, food, agriculture and manufacturing, and it represents the next great frontier for AI-driven value creation
- Most companies are applying AI as digital paint: adding it at the periphery of an existing industrial model without changing the core. The alternative — designing intelligence into the product and business model from the outset — is what Venkat calls digital structure
- Datagraphs, which connect entities, behaviours and time periods into a network of intent, are the industrial equivalent of what drove Netflix, Amazon and Spotify. Most industrial companies still hold their data in siloed systems of record
- The Nobel Prize in Chemistry going to Demis Hassabis, an AI researcher, is not an anomaly. It is evidence of the fusion that defines the intelligence age: traditional industry boundaries and AI are already converging
- The shift from economies of scale to economies of expertise means that a mid-sized chemicals or food ingredients company, if it has deep domain knowledge and the right ecosystem relationships, may be better positioned than it thinks
- The 90-day CEO plan is not about automating today’s processes. It is about understanding what your industry will look like in 2030 and working backwards from there
Moving Beyond the AI Fairy Dust
When I survey the executives I work with in agriculture, food and chemicals, the most common frustration I hear is not about a lack of ambition. It is about a lack of clarity. People understand, in broad terms, that AI is going to change things. They can see what it has done for asset-light businesses — the Netflixes, Amazons and Spotifys of the world. What they cannot see is what it means for a company that makes polymers, or catalysts, or crop protection chemicals.
Venkat Venkatraman has spent years building the framework that makes this legible. His book Fusion Strategy, published by Harvard Business Review Press, is one of the clearest accounts I have read of what the intelligence age actually means for industrial companies. It is not technical. It is not theoretical. It is a working model for leaders who need to act.
The starting point is the distinction between what Venkat calls digital paint and digital structure. Digital paint is when you take the existing industrial model as given and ask where AI can be added at the periphery — a better dashboard, a smarter email, an automated process. It is easy to do, easy to measure, and unlikely to change anything fundamental.
Digital structure is different. It asks: if intelligence were built into the core of this product or this business model from the outset, what would it look like? Tesla did not add digital features to a combustion engine. It designed a car as computers on wheels connected to the cloud and then built the physical manifestation around that. The distinction is not incremental. It is architectural.
The $75 Trillion Opportunity
The last two decades of digital transformation largely bypassed the physical world. The asset-light economy — software, media, financial services, digital platforms — was transformed first because data was already digital and products could be delivered as apps inside a smartphone. The camera, the music player, the payment terminal, the travel agency: all became software.
The asset-heavy world did not follow, not because transformation was impossible, but because the infrastructure to make it viable was not yet in place. The cost of sensors, satellite connectivity, cloud computing and analytics has now dropped to the point where it is.
Venkat’s rough calculation is that roughly $75 trillion of global GDP sits in asset-heavy industries that have not yet been systematically digitised. This is the territory where chemicals companies sell polymers without knowing how their customers use them. Where food ingredient suppliers have no visibility into the conditions under which their products perform. Where farmers buy inputs from multiple suppliers, none of whom can see the full picture of what is happening on the farm.
The same intelligence that allowed Netflix to predict what you want to watch before you know yourself can, in principle, allow a speciality chemicals company to predict when a customer’s process is about to fail before the customer’s own engineers detect it. The mechanism is identical. The scale of the untapped opportunity is vastly larger.
Datagraphs and the End of Systems of Record
The concept that unlocks this opportunity is the datagraph. Most industrial companies understand data as a system of record: a set of stored facts about what happened. Sales volumes, defect rates, customer lifetime values. Useful, but essentially backward-looking.
A datagraph connects those records into a network of relationships: between products and customers, between conditions and outcomes, between one customer’s behaviour and another’s. The result is not a description of what happened but a model of intent — one capable of predicting what is about to happen.
Netflix does not know what you want to watch tonight because it stores data about your past choices. It knows because it has built a graph that connects your choices to those of millions of other viewers across multiple contexts and time periods, identified the patterns that predict intent, and tested those predictions continuously at scale.
The adhesives supplier who instruments their customer’s production environment, measures temperature and humidity, and predicts drying times before the customer’s own engineers detect a problem is doing exactly the same thing. The technology is available. The bottleneck, as Venkat points out, is not data collection. It is two things: the analytical capability to turn connected data into actionable insight, and the customer trust required to earn the right to hold and use that data. These are business problems, not technology problems
John Deere and the Engaged Acre
The clearest example of an industrial company on this journey is John Deere. The concept that drives their transformation is the engaged acre: every acre of farmland on which John Deere equipment is operating and collecting data. The vision is not to sell more tractors. It is to accumulate a data graph of the farm, combining field performance, weather, agronomy, soil conditions and seed data, that gives John Deere a vantage point no individual farmer can match.
A farmer knows their own farm. John Deere, with data across thousands of farms in similar conditions, can see patterns the farmer cannot see, predict outcomes the farmer cannot predict, and recommend interventions the farmer cannot derive from their own experience alone. That is the data network effect: the value of the service increases with every additional farm that contributes to the graph.
This is not inevitable, and it is not guaranteed. John Deere, Cargill, BASF, and others are all competing to become the trusted orchestrator of the precision agriculture ecosystem. Who wins will depend not on who has the best equipment but on who earns the farmer’s trust to hold and optimise their data.
The Mindset Challenge
Venkat is clear that the constraint on this transformation is not technology. It is mindset. There are two specific shifts required.
The first is accepting that the future is not an extrapolation of the past. The core competencies that made industrial companies successful in the industrial age are not automatically the competencies that will make them successful in the intelligence age. Chemistry plus AI is a different discipline from chemistry alone. The Nobel Prize in Chemistry going to Demis Hassabis, the AI researcher who built AlphaFold, is not a curiosity. It is a signal.
The second is managing what Venkat calls the dual tension: running the existing industrial model at full performance while simultaneously designing the intelligence-age model that will eventually replace or transform it. This is harder than a technology upgrade. It is a business model transition — bigger, as Venkat and I agreed during the conversation, than the shift from steam to electricity that Joseph Fuller discusses in his episode on the steam engine mistake.
Economies of Expertise and the Mid-Sized Opportunity
One of the most energising parts of the conversation for me was Venkat’s argument about mid-sized companies. The examples we discussed, John Deere, Siemens, Novartis, are all multi-billion dollar organisations. The natural question for a CEO running a $300-500 million chemicals or food ingredients business is whether any of this is relevant to them.
Venkat’s answer is that mid-sized companies may actually have an advantage, for two reasons. First, private or closely held companies are not subject to the quarterly earnings pressure that forces large public companies to prioritise near-term profitability over long-term transformation. Second, the intelligence age is not primarily about economies of scale. It is about economies of expertise. The orchestrator of an ecosystem is the participant with the deepest and most distinctive domain knowledge, not necessarily the largest balance sheet.
A mid-sized speciality chemicals company prepared to move from macro to micro segmentation, to instrument its customers’ processes, and to share the resulting intelligence back through the value chain, is not at a disadvantage against larger rivals. It may be exactly the kind of high-expertise, high-trust partner that the ecosystem needs at its centre.
The 90-Day Plan
Venkat’s practical advice for a CEO wanting to act on this is deliberately restrained. Do not spend 90 days automating today’s processes. Spend them understanding what your industry will look like in 2030, which is only four years away, and whether your current model will still generate profit as well as revenue.
The method is scenario-based. Identify a maverick, someone in your organisation willing to argue that the current model will not survive, and let them make the case against someone defending the status quo. Use AI tools to stress-test both positions, not to generate the answers but to interrogate the assumptions. By the end of 30 days, you should have enough to convene a management conversation about which signals matter most and what experiments to run.
The point is not to predict the future accurately. It is to build the habit of looking at it honestly.
Listen to the Full Conversation
📺 Watch on YouTube: https://youtu.be/Vcb7e7M2UJU
🎧 Listen on Buzzsprout: https://www.buzzsprout.com/2012667/episodes/19324094
About the Guest
Professor Venkat Venkatraman is the David J. McGrath Jr Professor of Management at Boston University’s Questrom School of Business, where he holds the Chair in Management. He is one of the world’s leading authorities on how legacy industrial firms use real-time data networks and AI to reinvent their business models.
He is the co-author of Fusion Strategy (Harvard Business Review Press), available now, and the forthcoming Agentic Intelligence: Strategy at the Speed of Data (Forbes, 15 September 2026), which explores how humans and machines can work together to create intelligence.
Connect with Venkat on LinkedIn: www.linkedin.com/in/venkatraman/
Mark Blackwell is founder of Arkaro, specialising in change management, innovation, and commercial excellence for the agriculture, food, and chemicals industries. Arkaro’s collaborative “do it with you” approach works closely with clients through a four-step process: Understand, Co-create, Enable, Sustain. We don’t just coach — we get on the pitch with you.
Related Reading from Arkaro
On AI strategy and the intelligence age
The Steam Engine Mistake Companies Are Repeating with AI — Joseph Fuller, Harvard Business School — Joseph Fuller‘s argument that companies are bolting AI onto old processes rather than redesigning around it is the direct complement to Venkat’s digital paint vs digital structure framework. The steam engine metaphor that Mark references in the conversation comes from this episode.
Rewire or Retire: Why AI Is a Leadership Issue, Not a Technology Problem | Marco Ryan — Marco Ryan, referenced by Mark in the opening of this episode, on why AI transformation fails when it is treated as a technology project rather than a leadership challenge. The mindset argument runs parallel to Venkat’s.
Why AI Transformation Fails — and What Leaders Must Do in 90 Days | Charlene Li — Charlene Li on why most AI transformations fail because leaders are asking the wrong question — and what to do about it in 90 days. Venkat’s 90-day plan in this episode makes an interesting companion read.
AI and the Octopus Organization | Stephen Wunker — Stephen Wunker argues that AI will drive radical decentralisation of decision-making, requiring organisations to adopt a new anatomy: distributed intelligence with a strong central purpose. A direct complement to Venkat’s argument that the intelligence age demands a fundamentally different organisational model, not just a better technology stack.
On disruption and the pace of change
Why Smart Companies Miss Disruption: The 3 Ghosts Blocking Innovation | Scott Anthony — Scott D. Anthony on the organisational patterns that prevent established companies from responding to disruption. The companion piece to Venkat’s argument that the constraint is mindset, not toolset.
Why Playing Games at Work Isn’t Childish — It’s a Competitive Advantage | Scott Anthony — Scott Anthony’s argument that building the muscle for transformation requires deliberate practice, not just strategic planning. Connects to Venkat’s point that the 90-day plan is about building new thinking habits.
