An executive team prepares for a management review. AI has already summarized operating data, identified unusual patterns, compared current performance with history, extracted concerns from meeting transcripts, and produced several possible courses of action. Work that once took days to assemble is available before the meeting begins.

Management has more intelligence available to it than before. The more difficult question is whether that makes management more effective.

Herbert Simon and Peter Drucker approached this problem from different directions decades before today's AI systems existed. Simon examined decision making under the limits of human rationality and later considered what happens when organizations become information rich. Drucker concentrated on executive effectiveness: deciding what matters, focusing management effort, and producing results.

Neither was writing about AI, and neither should be presented as though he anticipated today's technology. Their work is useful for a different reason. AI changes several of the operating conditions they examined, allowing us to ask which management problems have been reduced, which have moved elsewhere, and which remain.

1. Information abundance creates a different scarcity

Simon explicitly examined the organizational consequences of an information-rich world. His argument went beyond the familiar idea of information overload. Information consumes the attention of its recipients. As information becomes abundant, attention becomes the scarce resource that organizations must conserve and allocate.

This was also an organizational-design problem. Simon considered how information systems and organizational mechanisms could filter and condense information rather than simply increase the amount presented to managers.

That distinction is particularly relevant under AI conditions.

AI lowers the cost of producing, retrieving, combining, and interpreting information. Management can compare evidence across systems, examine large amounts of unstructured material, retrieve historical information, and develop alternative interpretations much faster than before.

What does not expand at the same rate is management attention.

Simon therefore helps us see something important about AI-enabled management: abundance does not remove scarcity from the system. It can relocate it.

As producing analysis becomes less expensive, the management challenge increasingly shifts toward deciding where scarce attention and judgment should be applied.

2. AI changes some bounds, not the need for judgment

Simon's work on bounded rationality challenged models of decision making that assumed complete knowledge and unlimited capacity to calculate. His organizational work instead started from limits in human comprehension and computation under complexity and uncertainty.

Simon described division of decision work among specialists, coordinated through communication and authority structures, as one way organizations cope with bounded rationality.

AI changes some of those conditions substantially.

Managers can consider more evidence. Alternatives can be developed rapidly. Information distributed across documents and systems can be retrieved and compared. Computational limits that previously shaped decision preparation may become considerably less restrictive.

That does not mean bounded rationality disappears.

Some boundaries loosen while others remain. Management still operates with finite time and attention. Context may remain incomplete or contradictory. Future outcomes remain uncertain. Organizations have limited capacity to absorb consequences, and decisions continue to interact with authority, incentives, dependencies, regulation, customers, and other elements of the operating system.

There is also an additional judgment problem. Managers increasingly have to evaluate not only the operating evidence, but also the intelligence produced from it. A model may identify a strong pattern while an expert interprets the evidence differently. Historical experience may suggest one course while current conditions undermine an assumption on which that experience depended.

AI can expand the amount of evidence and analysis available to management. It does not eliminate the need to decide what is material, reconcile competing interpretations, or determine what organizational consequence should follow.

AI changes where some of the bounds are experienced.

3. Effectiveness still means concentrating on what matters

Drucker approached the management problem from a different direction.

His recurring concern was executive effectiveness. In The Effective Decision, he argued that effective executives concentrate on the important decisions, seek to understand the underlying realities, and focus on impact rather than technique. He was also skeptical of treating speed itself as evidence of effective decision making.

His later description of the effective executive follows the same pattern: asking what needs to be done, asking what is right for the enterprise, and taking responsibility for decisions and their implementation.

Drucker's broader distinction between efficiency and effectiveness reinforces the point. Doing work efficiently is not the same as directing effort toward the right work.

Viewed under AI conditions, this creates an important question:

If producing analysis becomes dramatically easier, does the organization become more effective, or merely more efficient at producing intelligence?

AI can improve a forecast without determining whether the forecast addresses the decision that matters most. It can improve portfolio reporting without determining where executive intervention will create the greatest contribution. It can produce increasingly sophisticated recommendations while management continues to focus attention on issues with limited connection to enterprise performance.

Drucker's skepticism about decision speed is also useful when AI makes analytical speed highly visible. Faster preparation can remove genuine delay. But effectiveness still depends on whether management is addressing the right question, making the required decision, and producing an appropriate organizational result.

AI improves the means available to management. Drucker's effectiveness question keeps attention on the purpose.

4. What Simon and Drucker reveal under AI conditions

Taken together, Simon and Drucker provide a useful way to interpret the management consequences of abundant AI-generated intelligence.

Simon shows why increased information does not remove scarcity. As information becomes easier to produce and distribute, the capacity to give it meaningful attention remains constrained.

Drucker reminds us that management effectiveness is not determined by how much information is processed or how much analytical work is produced. Management has to concentrate on the work that matters and translate knowledge into contribution and results.

Viewed through today's AI conditions, the production of intelligence can become substantially cheaper while the organizational resources required to convert intelligence into consequence remain constrained.

Attention still has to be allocated. Judgment still has to be exercised. Decisions still require legitimate authority. Action still has to be coordinated. Results still have to be connected to organizational purpose and enterprise value.

An organization can therefore possess more AI capability without possessing proportionately greater management capability.

It may become more observable. It may produce better analysis. It may reduce the effort required to develop recommendations. Those are meaningful improvements, but they still have to enter a management system capable of deciding what deserves attention and what contribution is required.

The advantage increasingly lies not in access to intelligence alone, but in the organization's ability to use abundant intelligence selectively and effectively.

5. The AIMR extension

This is where established management thinking meets an operating condition neither Simon nor Drucker had reason to address directly.

AI-enabled systems can increasingly be used across much more of the management-information chain: sensing operating conditions, assembling evidence, retrieving retained context, identifying patterns, developing alternatives, generating recommendations, monitoring assumptions, and, in some settings, participating directly in execution.

AIMR's extension is to examine what the surrounding management system must do when those capabilities become widely available.

The challenge is not simply to provide managers with more information. It is to convert available intelligence into sufficient organizational understanding for management to determine what matters, allocate attention appropriately, exercise judgment, make legitimate decisions, and translate those decisions into coordinated action.

That makes the interfaces between intelligence and management more consequential.

Analytical capability may scale rapidly while management attention does not. Recommendations may arrive before appropriate authority is available. Current information may become readily accessible while the assumptions behind earlier decisions remain poorly retained. Local AI capability may improve without those improvements combining into enterprise performance.

These are AIMR extensions, not propositions that Simon or Drucker made about AI. Their work provides the intellectual starting point.

Simon helps explain why information abundance does not abolish scarcity. Drucker helps explain why producing more knowledge does not by itself create effectiveness. Under AI conditions, those ideas converge on a management requirement: organizations must become better at determining where intelligence deserves attention and how it should be converted into purposeful, legitimate action.

When intelligence becomes abundant, management effectiveness depends less on producing more information and more on determining what matters and converting that understanding into effective action.

Sources and further reading

• Herbert A. Simon, Designing Organizations for an Information-Rich World (1971).

• Herbert A. Simon, Rational Decision-Making in Business Organizations, Nobel lecture (1978).

• Peter F. Drucker, The Effective Decision, Harvard Business Review (1967).

• Peter F. Drucker, What Makes an Effective Executive, Harvard Business Review (2004).

• Peter F. Drucker, The Effective Executive.

• Drucker Institute, source archive on Drucker's distinction between efficiency and effectiveness.