The New Scarcity Is Reliable Access

MGF Weekly Observation | July 13–19, 2026 (W29)

During the third week of July 2026, several developments unfolded at the same time.

Military exchanges between the United States and Iran continued. Maritime transport and energy security became more uncertain. Oil prices rose, while financial markets once again focused on inflation and interest-rate risk.

During the same week, the global AI competition also shifted in character. AI was no longer being treated only as a product category in which companies compete over model quality. It was increasingly being organized as strategic infrastructure governed through semiconductors, computing capacity, electricity, export controls, standards, and international partnerships.

War, oil, finance, AI, and electricity may appear to be separate stories.

Structurally, however, they are converging on the same physical foundations:

• energy
• power grids
• semiconductors
• maritime routes
• capital
• industrial capacity

The central change visible in Week 29 was not simply that the world was running short of resources.

It was that access to critical resources and infrastructure was becoming more conditional, more closely monitored, and more unevenly distributed.

---

War Changes the Conditions of Passage

The effect of military conflict on logistics is not limited to destroyed ships, damaged ports, or a formally closed waterway.

During Week 29, repeated military action and maritime restrictions increased uncertainty around shipping routes connected to the Gulf.

A vessel’s ability to pass through a contested region is not determined by a simple yes-or-no decision. Commercial access depends on several overlapping conditions:

• flag and registration
• beneficial ownership
• cargo
• previous port calls
• insurance coverage
• sanctions exposure
• access to payment and settlement systems

A route can remain physically open while becoming commercially, legally, or operationally difficult to use.

A ship may still be able to sail through a strait, but that does not mean it can do so safely, on schedule, under an acceptable insurance policy, and at a viable cost.

Military pressure can therefore weaken supply before it physically reduces the volume of goods moving through the system.

What changes first is the reliability of connection.

This is the beginning of probabilistic logistics: a condition in which goods continue to move, but the confidence that they will arrive on time, at a predictable cost, and under stable legal conditions begins to decline.

---

Oil Prices Are Also Prices of Future Connectivity

Oil prices do not reflect only the amount of oil currently available.

They also absorb expectations about whether the system will remain connected.

Markets evaluate questions such as:

• Will major shipping lanes remain usable?
• Will insurers continue to cover voyages through the region?
• Will military operations expand?
• Will neighboring states become more directly involved?
• Will alternative routes remain economically viable?

For this reason, oil prices can rise even when an immediate physical shortage has not yet occurred.

The market is not only asking how much oil exists.

It is asking whether that oil can continue to move through the same routes, under the same contracts, with the same level of confidence.

Oil prices are therefore both a measure of physical supply and a measure of expected connectivity.

Week 29 did not prove that energy flows had collapsed.

It showed that confidence in those flows had weakened.

---

AI Is Becoming a System of Infrastructure Governance

The same structural shift can be seen in AI.

Until recently, much of the public discussion focused on model performance, user numbers, semiconductor speed, and corporate valuation.

Those measures remain important, but a different set of questions is becoming more central:

• Who can obtain advanced chips?
• Which countries receive access to large-scale computing capacity?
• Where are data centers allowed and encouraged to expand?
• Which cloud platforms and technical standards become dominant?
• Who controls exports, licensing, and international cooperation?

AI is increasingly shaped by infrastructure governance.

This does not mean that an AI model itself becomes a national operating system.

It means that states are beginning to arrange and govern the systems around AI at a national or geopolitical scale:

• semiconductors
• computing capacity
• electricity
• cloud infrastructure
• technical standards
• export controls
• international partnerships

China has sought to widen its AI cooperation networks and build connections across emerging economies.

The United States has continued to protect its position in high-end chips and upstream computing infrastructure through access controls and export rules.

Their strategies differ, but both increasingly treat AI as a strategic resource whose conditions of access can be politically managed.

The competition is shifting from:

Who builds the best model?

toward:

Who controls the infrastructure that determines where advanced AI can be built, operated, and expanded?

---

Fast AI Depends on Slow Physical Systems

AI models, market valuations, export regulations, and national policy announcements can change within days.

Physical infrastructure cannot.

The following systems require years of planning, construction, negotiation, and maintenance:

• power generation
• transmission grids
• data centers
• semiconductor fabrication plants
• cooling systems
• water infrastructure
• workforce development

This creates a major velocity mismatch.

Fast layer:
AI models, finance, policy, regulation

Slow layer:
electricity, construction, logistics, industrial capacity, local communities

When fast-moving systems pull on slow-moving infrastructure, friction does not necessarily appear inside AI itself.

It appears at the boundaries:

• higher electricity prices
• grid-connection delays
• construction bottlenecks
• water demand
• backup-generation requirements
• local infrastructure costs
• disputes over who pays for expansion

The faster AI investment grows, the more important these slower physical constraints become.

---

Constraints May Concentrate Growth Rather Than Stop It

Higher energy prices and higher capital costs do not automatically bring AI investment to a halt.

They may instead determine which actors are still able to participate.

Large corporations and states can absorb constraints through combinations of:

• long-term financing
• dedicated power contracts
• public subsidies
• semiconductor purchasing power
• legal capacity
• insurance
• political and commercial bargaining power

Smaller firms, new entrants, and resource-poor regions cannot secure the same conditions.

The result may not be a smaller market.

It may be a more concentrated one.

Resource constraint
↓
Higher cost of participation
↓
Fewer actors able to absorb uncertainty
↓
Greater concentration among states and large corporations

In this structure, constraints do not simply slow growth.

They select who is allowed to continue growing.

---

Stability in the Gulf Is Being Produced Through Compensation

States surrounding a conflict zone may not be direct combatants, yet they absorb several forms of pressure.

These can include:

• interception and air defense
• military-base operations
• airspace restrictions
• port management
• market losses
• rerouted flights and shipping
• adjustments to energy-export routes

A city or market may continue to function and therefore appear stable.

But stability should not be measured only by whether activity has stopped.

It should also be measured by what must be done to keep activity running.

If normality depends on interception systems, emergency routing, reserves, insurance adjustments, administrative coordination, and continuous local problem-solving, then the system is not operating without pressure.

It is operating through compensation.

The longer compensation continues, the more likely it becomes that costs move beyond the military sphere and into finance, logistics, public infrastructure, and household life.

---

The Household Layer Is Still Quiet

During Week 29, the effects of oil prices, shipping risk, interest rates, and AI infrastructure spending had not yet fully appeared as a large public reaction at the household level.

That silence should not be interpreted as proof that no pressure exists.

There is a delay between market-level change and everyday cost.

Oil, currency, and insurance
↓
Imports, wholesale markets, and logistics contracts
↓
Corporate costs
↓
Consumer prices and utility bills
↓
Household life

The system was still moving through this transmission path.

A similar silence surrounds AI infrastructure.

National growth, corporate investment, and technological leadership are widely discussed.

The distribution of costs is discussed much less clearly:

• grid upgrades
• electricity prices
• water use
• construction costs
• local infrastructure burdens
• delays in connecting other users to the grid

This silence contains several possible conditions:

• the cost has not yet reached households
• the cost has arrived, but its cause is difficult to identify
• the cost has been distributed across many separate contracts and bills
• public language for describing the burden has not yet formed

Silence does not necessarily mean the absence of pressure.

It can mean that pressure has not yet been translated into everyday language.

---

The Global System Is Not Weakening in the Same Way Everywhere

It would be misleading to describe the current situation simply as global instability or collapse.

Different layers are changing in different directions.

At the center, states and large corporations are becoming more capable of setting access conditions.

In the middle, finance, insurance, logistics, ports, utilities, and public administration are absorbing uncertainty.

At the periphery, regions, smaller firms, and households receive the costs later.

Center:
sets the conditions of access

Middle:
translates and absorbs uncertainty

Periphery:
receives the cost with a delay

The global system is not becoming uniformly thinner.

The areas that control access are becoming more rigid.

The areas that depend on access are becoming more exposed.

The inequality emerging here is not only a difference in how much a country, firm, or household owns.

It is also a difference in whether they can:

• obtain access
• preserve access
• absorb interruptions
• remain connected during exceptional conditions

This is a form of inequality based on continuity of access.

---

What Changed in Week 29

The week of July 13–19 did not show that global systems had stopped functioning.

Oil continued to flow.

Ships continued to move.

AI investment continued.

Markets continued to operate.

But the nature of connection was changing.

Reliable access
↓
Conditional access
↓
Monitored access
↓
Probabilistic access

The central question is no longer only whether a resource exists.

It is whether a particular actor can reach it, finance it, insure it, and remain connected to it under changing conditions.

These conditions are increasingly being reorganized by states, military power, financial systems, and large corporations.

The new scarcity is not only energy, computing power, capital, or transport capacity.

The new scarcity is the ability to remain reliably connected to them.

Week 29 was not simply a week in which the world moved closer to shortage.

It was a week in which the global system became more selective about who could stay connected.

---

Branch Gradient Log

Dominant conditions

Military action and maritime monitoring continue. Oil, insurance, and logistics costs remain elevated. State control over AI and semiconductor access expands. Actors with sufficient capital, energy access, industrial capacity, and political support gain a larger advantage.

Reversal conditions

Military operations decline and shipping conditions become verifiable. Insurance and transit rules become clearer. International coordination around AI access improves. The cost of data-center expansion and grid development becomes more transparent and institutionally distributed. More decentralized routes to energy and computing capacity become operational.

Current gradient: Strong

Fast-moving military, financial, and AI systems are converging on slow-moving energy, logistics, and electricity systems.

At present, the stronger gradient points not toward a complete stop, but toward stricter access conditions and greater concentration among actors able to absorb the constraints.

---

Translation Layer | Decision Contact and Recursive Review

Decision contact

This structure becomes relevant wherever governments balance security against economic continuity, companies plan around long-term access to power and computing capacity, investors evaluate operational resilience rather than growth alone, and institutions determine how infrastructure costs are distributed between national beneficiaries and local communities.

Recursive review

The structure depends on slow physical systems continuing to absorb decisions made by faster military, financial, and technological layers. The phase changes when insurance, grid capacity, capital costs, or regional burden reaches a compensation limit. Variables requiring renewed observation include actual shipping activity, insurance conditions, grid-connection delays, infrastructure concentration, and the lag between upstream costs and household prices.