The Sustainable Machine: Why Digital Transformation Must be Designed for Resilience

Senior leaders are being asked to approve increasingly complex investments in artificial intelligence, blockchain and the Internet of Things (IOT). The challenge is no longer deciding whether these technologies matter but determining whether they will create durable value or introduce new forms of cost, dependency and risk. They are routinely presented as tools that will make organizations more efficient, economies more transparent and societies more sustainable.

They can achieve all three. Yet they do so far less reliably than most commercial accounts suggest. For boards and executive teams, understanding the gap between technological promise and practical results is becoming an essential part of investment, governance and risk management.

Across research institutions, government advisory work and large firms, we repeatedly encountered organizations that had invested in credible technology, hired capable people and committed substantial budgets, only to find that the expected value failed to materialize. For senior leaders, there must be a critical distinction between funding a technical deployment and enabling genuine transformation.

Why Technology Alone Does Not Create Value

Value does not reside inside an algorithm, a sensor or a distributed ledger. It emerges from the relationships between technologies, organizations, incentives and people. For that value to materialize, people must be both able and willing to act on what a system tells them. Technology investment cannot be assessed separately from operating models, decision rights and organizational behaviour.

This is why we ask one question of every technology proposal we are shown: what does it converge with? In practical terms, what other systems, processes, institutions and people must it work with to create a useful capability? A sensor feeding a dashboard that nobody opens is telemetry, not transformation. A machine learning model without timely operational data is a demonstration. A distributed ledger operated by a single trusted organization is little more than a slow and expensive database.

When convergence is designed well, each technology has a clearly defined role. Connected devices observe events in the physical world, AI systems interpret patterns in those observations, cloud infrastructure distributes data and computing power and distributed ledgers can maintain a shared record between organizations that do not wish to depend on a single central intermediary. The strategic value lies in the capability created by the combination, not in any individual component.

However, convergence is not automatically benign and this is the part that is often overlooked in investment cases. A disconnected sensor is merely ineffective, while a connected network of poorly secured sensors is an operational and security liability. An isolated algorithm has limited reach, but an algorithm embedded in thousands of decisions can reproduce its errors at scale.

Leaders should therefore ask not simply whether technologies are connected, but what kind of system their connection creates, who benefits from it and who carries the operational, financial and reputational risk when it fails.

The Full Cost of Digital Systems: AI and Data

Digital systems are routinely described through metaphors of weightlessness: data moves through “clouds”, while AI appears to inhabit an abstract world of models and mathematics. These metaphors can make the material foundations, operating costs and long-term dependencies of digital technology easy to overlook.

In practice, every digital operation depends on physical infrastructure, including mines, factories, cables, data centres, cooling systems, electricity grids and discarded devices. The cloud is not immaterial. It is an industrial landscape viewed from a distance and its costs and risks remain part of the organization’s value chain even when they sit outside its direct control.

Efficiency also deserves scrutiny. A more efficient AI model may consume less energy for each task while encouraging an organization to run many more tasks. Efficiency is worth pursuing, but it does not automatically reduce total environmental impact. In some cases, it simply makes higher levels of consumption more affordable. A credible business case must therefore consider total demand, not only the efficiency of each individual transaction.

The social and regulatory risks deserve the same degree of scrutiny. AI can identify patterns that people would miss, but it can also give an appearance of mathematical authority to patterns shaped by historical discrimination, incomplete data or careless assumptions. When AI influences employment, insurance, healthcare or access to public services, an error is not merely an inconvenience. It is a consequence borne by a person and, ultimately, a source of legal, reputational and operational exposure for the organization responsible.

Human oversight is the standard response, but placing a person somewhere in the process does not, by itself, create accountability. The real question is whether that person can understand the decision, challenge it and be answerable for its consequences. Frequently, they can do none of the three. Senior leaders should be wary of governance arrangements that appear reassuring on paper but provide no meaningful authority to intervene.

Digital Sustainability: Resilience as the Real Test

Digital sustainability is about more than carbon. For boards and executive teams, its real test is whether a system can absorb disruption without creating new and less visible forms of fragility. A system may be efficient under normal conditions yet become dangerously brittle when a supplier fails, a network is attacked or a regulatory requirement changes. Resilience therefore belongs within business continuity, procurement and strategic planning, rather than being treated as a purely technical concern.

The same trade-off appears at every layer. Cloud platforms can make individual organizations more resilient while making society dependent on a small number of providers. Blockchain can distribute authority while concentrating influence among developers and consortium members. AI can widen access to expertise while centralizing the data, energy and computing power required to build the most capable models. Each may strengthen an organization in one respect while creating a new concentration of risk in another.

No organization can eliminate all its dependencies, but it can make them visible and deliberate. The leadership task is to understand which dependencies are strategically acceptable and which leave the organization without meaningful alternatives. Open standards, data portability, systems built from replaceable parts and a properly costed migration plan all help to preserve strategic options.

An exit plan is not a declaration of intent to leave. It is what makes leaving possible if prices rise, service deteriorates, the technology is superseded or the law changes. Without one, a successful implementation can become a source of vendor lock-in and diminishing strategic control. Resilience is not only the capacity to keep operating. It is also the capacity to adapt.

Leadership Determines What AI Will Amplify

AI, blockchain and connected devices are neither saviours nor villains. They are amplifiers of human capability and like all amplifiers, they enlarge our wisdom and our short-sightedness, our capacity to cooperate and our capacity to exploit, without discriminating between them. The role of leadership is to determine what those systems will amplify and to ensure that the organization can answer for the consequences.

The question that matters for senior leaders is not how intelligent our machines become. It is whether the institutions that build, buy and govern them become capable of understanding, managing and answering for what they do.

Save 30% on Digital Sustainability with code ABS30.