Fusion Strategy will transform the way businesses are run

We used to talk about digital strategy in the past. That is about automating processes. Think: SAP, ERP, and other large-scale investments to automate processes. We used to automate supply chain, accounting and so on. That improved efficiency but it did not change strategy. It did not create new business models.

Fusion Strategy is about creating new business models and developing new strategies. It’s about digitising the product which, in turn, changes strategy. We argue that industrial companies must combine what they do best - create physical products - with what digitals do best - use AI to analyse large product in-use datasets - to develop new strategies. In other words, the laws of competitive advantage are changing, rewarding those who have the most robust real-time insights rather than the most valuable physical assets. Consider these company examples that highlight how, for asset-heavy industrial companies, competitive advantage is shifting to data and AI.

Rolls-Royce collects enormous amounts of data on every flight that uses its aircraft engines. Rolls-Royce analyses fuel consumption patterns based on product-in-use data such as the routes aircraft take, the altitudes at which they fly, the weather conditions during flight, the speed at which they fly, and the load the aircraftis carrying. Rolls-Royce receives more than 70 trillion data points each year from its engines. By leveraging the power of data and AI, Rolls-Royce helps airline companies to optimise fuel efficiency, estimating that improving an engine’s fuel efficiency by one per cent results in $2 billion of annual savings for an airline.

View, Inc. has developed smart glass, powered by data and AI, that automatically adjusts in response to the sun, increasing access to natural light while minimising heat and glare. View-designed windows provide more energy-efficient office buildings while avoiding the need for expensive window shades.

In the same industry, Corning could collect detailed images on the drop performance of its latest Gorilla glass used on different smartphones to train its foundation models to derive insights to design future products.

Fusion Strategy embodies five principles:

Integrate steel and silicon. Tractor is not just Big Iron but is about leveraging data and AI on the physical tractor to create new value.

Combine intelligent humans with powerful machines. AI should be called augmented intelligence, not artificial intelligence.

Infuse digital into analog disciplines of sciences, arts, and engineering. Historically, computer science was regarded as distinct from medicine, law and economics. Now every discipline is being impacted by digital. In health, the future is personalised health with biomarkers and customised cures.

Link the physical and virtual worlds through digital twins, mixed reality, and metaverses. We must simultaneously operate in the physical and virtual worlds, in designing and building a new factory.

Move away from a firm-centric view to an ecosystem view. Value is created by orchestrating ecosystems with multiple partners.

The auto industry has made the most progress in Fusion Strategy. Tesla is an example and it follows the five principles. Tesla can observe every mile their vehicles travel on roads using the multiple cameras designed into the car’s body. Based on product in-use data, Tesla can plan predictive maintenance on its cars through over-the-air software updates. Every Tesla is designed to interconnect the physical and digital domains, creating the ability to collect data in motion (fusion principle #1).

Every Tesla has an in-built “shadow mode” that simulates, even if the car’s autopilot computer is not in use, the driving process in parallel with the human driver. When the algorithm’s predictions don’t match the driver’s behavior, snapshots of the car’s cameras, speed, acceleration, and other parameters are recorded and uploaded. Tesla’s AI team reviews and analyses the data to identify human actions that the system should imitate and use as training data for its neural networks. For instance, they may notice that the system fails to identify road signs obscured by trees and figure out ways to get better-quality data.

Fusion Strategy happens when intelligent humans learn with powerful machines (fusion principle #2).

Winning in the automotive future calls for seamlessly integrating traditional competencies in automobile design and manufacture with emerging digital disciplines such as hardware, software, applications, connectivity, telematics, and analytics. Mercedes Benz and VW are committed to developing their operating systems and master software competencies. Cruise has prototyped Origin — a zero-emission, shared, electric vehicle designed from the ground up to operate without a human driver and is a reimagination of the automobile without human-centered features like a steering wheel or sun visor. Waymo has prototyped its vision of vehicles without steering wheels, accelerators, and brakes with Geely’s Zeekr. Automakers must become digital engineering companies with competencies at the intersection of traditional disciplines and digital technologies (fusion principle #3).

Metaverse is key to te automotive future. BMW, for instance, is using Nvidia’s Omniverse platform to build a new factory where people and robots work closely, and engineers collaborate in virtual spaces. With information from design and planning tools generating realistic images of the planned factory, BMW can assess the critical trade-offs it must make in production systems. In addition to factory design, Nvidia’s platform allows automakers to evaluate how autonomous vehicles perform on the road by generating simulations of highways or urban streets to test the vehicle’s perception systems, decision-making capabilities, and control logic (fusion principle #4).

Finally, automakers cannot create value by themselves but they must put together an ecosystem of partners. GM is scaling Cruise in partnership with Honda, Microsoft, and Walmart. GM’s Ultium battery and motor, developed with LG Chem, might invite other automobile manufacturers to become partners once it reaches production for deployment at the scale stage. Motional — the joint venture between Hyundai and Aptiv — has partnered with Uber for autonomous rides and delivery. Tesla, which has open-sourced its patents, could invite other automakers to use its Dojo supercomputing capacity for testing to enhance the reliability and safety of autonomous driving systems. Tesla’s charging stations are used by several automakers. Uber’s ability to coordinate and match riders with drivers at every one of the thousands of cities they operate at every moment of request is based on assembling the ecosystem of relevant partners and ensuring that they have real-time data to deliver the services (fusion principle #5).

Fusion Strategy requires complete overhaul of the organisation — talent base, its structure, training employees, process of creating and capturing value.

We expect the following fundamental changes:

Instead of linear and gradual growth, industrials will experience non-linear and exponential growth.

Instead of focusing on familiar competitors with similar business models, industrials will compete with digital natives with new capabilities.

Instead of production-based scale, industrials will shift to data-based scale.

Instead of product-market extensions, industrials will compete with data and AI to redefine industry boundaries.

Instead of vertical integration, industrials will pursue virtual integration.

Instead of firm-centric approach, industrials will compete in ecosystems.

Instead of siloed databases, industrial will integrate all product in-use data in one location that can serve the needs of all managers.

We are not arguing for Big Data. We are not arguing for volumes of data. In fact, most companies are drowning in data lakes, not knowing how to create value from data.

We argue for Smart Data. Smart Data is not about volume, but it is about value creating data. This is the idea behind product in-use data. Industrial companies collect data on products as sold. For instance, General Motors knows how many cars it sold. Smart Data is about all the data from the car that the customer of GM is driving. This is what we mean by product in use data.

Smart Data leads to intelligent strategic moves.

Gen AI is the next inflection point in the evolution of AI. It can generate new content in unstructured forms from foundation models from Open AI, Anthropic, and others. Gen AI chatbots leverage foundation models and extensive neural networks trained on large, diverse, quantitative, qualitative, and unstructured data sets, enabling various tasks. Unlike narrow AI, which performs single tasks like predicting customer churn or optimising production runs, Gen AI models are versatile and capable of tasks such as summarising technical reports, developing new product ideas, providing varied recipes, and doing complex programming.

While the headlines focus on Gen AI’s ability to write essays, author poems, and create images, melodies, and movies, its real benefits will be its ability to transform business logic, create new sources of competitive advantage, and render traditional competencies obsolete. Gen AI isn’t just about incremental productivity improvements; it will create new forms of economic value. In the process, it will reshape the nature of competitive interactions in industries and ecosystems.

The Internet allowed companies to create e-Commerce channels, while smartphones enabled m-commerce. Those two innovations predominantly influenced consumer settings. Gen AI will likely impact industrial companies. With the technology able to, among other things, generate complex designs; extract insights and trends from multi-modal data; predict and respond proactively to changing conditions; and handle ambiguous and incomplete data, Gen AI is tailormade to transform the logic of competition in industrial businesses. GAI can answer more complex questions and solve non-linear problems accurately and quickly when trained with appropriate context-specific data. Multiple studies — although preliminary — highlight that the areas where Gen AI will make the most impact in the next 18 months will be in functions and sectors that are asset-heavy and information-rich.

I do not think so. AI is a promising technology with benefits across sectors including education, health, and renewable energy. All the same, there are three major risks with AI:

Job losses: AI will not obsolete jobs but people who use AI will obsolete the jobs those that don’t use AIU. Implication is that we need large-scale retraining of the workforce.

Malicious use: AI creates lots of autonomous agents. For instance, financial systems could be run by AI. In such a scenario, rogue players can hack into the system and cause significant damage to the economic system. It is, therefore, critical that regulations and guardrails are in place to prevent such abuse.

Existential risk: This is the extreme case where AI becomes so powerful that it destroys humanity. Sort of what intelligent machines try to do in movies like the Terminator or Matrix. This is possible with General Artificial Intelligence (GAI). We are far away from developing GAI. I personally think that the existential risk posed by AI is low.

You quote Elon Musk regarding the dangers of AI. If he really believes in that, why has he started an AI company, xAI? Open AI was created to develop AI in an extremely methodical and slow way. If that is their original mission, why have they taken billions of dollars from a commercial enterprise like Microsoft. Anthropic was founded by a few who left Open AI violently disagreeing with Microsoft’s investments in Open AI. If so, why has Anthropic accepted $4 billion investment by Amazon?

My broader point is that the AI revolution that we have witnessed in the past 18 months is good for humanity. It will bring a lot of benefits, especially for countries like India. Why? In the West, we worry that AI will disrupt white-collar jobs. In India, we have a shortage of white-collar professionals. We have few doctors, few surgeons, few schoolteachers, few college professors, and so on. India should embrace and leverage AI to dramatically improve productivity of the scarce resource, white-collar professionals.