Cheng Maiyue: The United States doesn't lack electricity, but rather its power grid.

(Image caption) With AI, military research and development, and key industries heavily reliant on electricity, grid resilience is gradually rising to the level of a national security issue. The power system is no longer just an engineering problem for the energy sector, but a strategic infrastructure affecting institutional efficiency, industrial competitiveness, and security boundaries.

 

 

The article, "Power Shortages: A Hidden Constraint on U.S. National Security," is one of the key research texts in the GFM "AI Energy" special issue.

 

Before its official publication, GFM, following its established research and editing procedures, proactively provided the full text to Mr. Cheng Maiyue, the column's host and advisor, an expert in energy and power systems, for professional pre-publication calibration and review.

 

After reading the entire article, Mr. Cheng Maiyue agreed with the research questions and direction set forth in it. Based on this, and through professional discussions with the editor-in-chief, he offered suggestions for correction and supplementation regarding the use of certain terminology, causal chain expressions, and engineering logic. These revisions have been faithfully reflected in the article without altering the original research stance and narrative structure.

 

This article is compiled based on discussions conducted by experts invited by the editor-in-chief of GFM during the research and writing phase, aiming to accurately present their engineering judgments, analytical approaches, and feasible paths. The macro-level inferences presented herein are presented as expert opinions and do not constitute any policy recommendations, investment advice, or endorsement of any position.

This "Editor's Note" is the only editor's statement of responsibility in the entire text.

 

⸻

 

Editor's Note | Why we chose to calibrate before publication

 

During our research for the GFM "AI Energy" special issue, we gradually confirmed that a structural shift is underway:

As artificial intelligence, data center construction, military research and development, and energy transition accelerate simultaneously, electricity issues are no longer just supporting conditions at the industrial level, but are gradually evolving into key constraints affecting national capabilities, institutional efficiency, and systemic resilience.

 

The research text, "Power Shortages: A Hidden Constraint on U.S. National Security," was completed against this backdrop.

However, before formal publication, GFM chose to put the research content under engineering and system-level testing first, and through the pre-editing calibration mechanism, confirmed whether the problem breakdown and causal logic were valid.

 

This process is not about obtaining any form of endorsement, but rather about returning to two of the most fundamental, yet most easily overlooked, questions in research writing:

 

(Image caption) Most of the high-voltage transmission networks in the United States were built in the mid-20th century. Long-term underinvestment and disrepair have become core bottlenecks restricting the integration of new energy sources and power dispatch. The root cause of the power problem is not insufficient power generation capacity, but rather the inability of the transmission and distribution system to handle rapidly growing loads and the demands of energy transition.

 

 

For GFM, "AI Energy" is not a column that prioritizes commentary, but rather a research project that attempts to examine technological realities, engineering constraints, and institutional judgments within the same framework. Under this premise, pre-editing calibration is not an additional procedure, but a necessary condition for the validity of the content.

 

⸻

 

01 | First level of calibration: not adjusting stance, but correcting the way of expression

 

During the review process, Mr. Cheng Maiyue expressed his agreement with the article's problem awareness and overall research direction; at the same time, he also pointed out that if such research is to have the possibility of being cited by higher levels or even across departments in the future, certain expressions that are easy to cause misunderstanding in professional contexts must be avoided.

 

The most direct aspect comes from the calibration of terminology context.

 

For example, the abbreviation HVDC that originally appeared in the text may have different technical meanings in different engineering scenarios. Cheng Maiyue's suggestion is very clear: in such research texts, the unambiguous term "high-voltage direct current transmission" can be used directly, without the need to emphasize the abbreviation.

 

The reasons are quite clear:

Professional readers may become confused by the context, while general readers have no actual need to understand it.

 

⸻

 

02 | Key Calibration: The problem isn't "not generating electricity," but "not being able to send it in."

 

When discussing issues such as renewable energy growth, project delays, or partial power phase-out, Cheng Maiyue repeatedly reminded:

The above situations are mostly "superficial phenomena" in system operation, rather than structural root causes.

 

(Image caption) High-voltage direct current (HVDC) transmission is an engineering technology for long-distance, high-capacity power transmission, not just a conceptual slogan. Its deployment is highly dependent on cost, planning, and institutional conditions, and must be used prudently within the correct engineering context.

 

 

From the perspective of power system engineering, the core challenge currently facing the United States is not a lack of capacity to build power generation facilities, but mainly manifested in the following aspects:

 

This is why many new energy projects, even after completion, still have to wait a long time to be connected to the grid, making it difficult to convert them into usable power resources immediately.

 

⸻

 

03｜Two engineering analogies to illustrate that "the power grid is the bottleneck"

 

To avoid the concept of "grid capacity" remaining merely an abstract one, Cheng Maiyue proposed two engineering analogies during the discussion.

 

highway metaphor

Power generation capacity is like the supply of vehicles, while the power grid is like the highway. Vehicles can continue to increase, but if the roads are not expanded in the long term, the result will not be a shortage of vehicles, but rather a limitation on transportation capacity.

 

Weightlifting metaphor

If a system operates near its limit for an extended period, its stability will naturally decline. Only with sufficient margin can a system remain safe and resilient when demand increases.

 

These two metaphors point to the same conclusion:

The power grid's reserve margin and expansion capacity directly determine whether it can meet the rapidly growing electricity demand in the AI era.

 

 

 

(Image caption) While the installed capacity of wind and solar power in the United States continues to climb, a large amount of electricity is forced to be curtailed or abandoned due to insufficient transmission and grid connection capacity. The problem with new energy sources is not that there is "not enough," but that they "cannot be transmitted," highlighting the structural contradiction of lagging grid investment and planning.

 

 

04 | Word Choice Recalibration: From "Reliable" to "Schedulable"

 

In describing the power supply structure, Cheng Maiyue also proposed a seemingly minor but actually crucial correction.

 

Compared to the more narrative-driven term "reliable power," he suggests using the more engineering-specific term "dispatchable power."

This adjustment allows the discussion to directly address core issues such as system scheduling, peak-valley balancing, and capacity management, rather than merely focusing on the intuitive level of "reliability."

 

⸻

 

05 | Engineering Breakdown Methods: Why Can We Only Solve "Local Loads" First in the Short Term?

 

After clarifying the root causes, Cheng Maiyue further proposed an engineering solution that is in line with the actual conditions.

 

Its basic judgment is as follows:

On the one hand, the deployment speed of photovoltaic and energy storage systems is significantly faster than the overall upgrade of the power grid;

On the other hand, the acceptance, expansion and management of the power grid often require a longer engineering and institutional cycle.

 

Therefore, given the rapid growth in electricity demand from AI and data centers, a more feasible approach in the short term is to deploy power generation and storage facilities close to the load end (such as data centers) to directly supply electricity, thereby bypassing the bottleneck of the traditional power grid and forming a local power supply closed loop.

 

This approach does not negate the power grid itself, but rather represents a phased choice based on engineering realities:

Overhauling the power grid involves right-of-way, approvals, and cross-level governance, making it difficult to keep pace with the demand curve of AI.

 

⸻

 

06 | Why can't we wait for "the whole system to be slowly fixed"?

 

Cheng Maiyue specifically pointed out that the reason why the United States has gradually formed the expectation of "long-term power shortage" is not due to the stagnation of power generation capacity, but because the engineering and governance cycle for improving power supply capacity has obviously lagged behind the speed of demand growth.

 

(Image caption) The microgrid model, which combines photovoltaics, energy storage, and data centers, can bypass traditional grid bottlenecks and address high-load electricity demand at specific points in certain scenarios. However, this type of solution is a localized approach and cannot replace a nationwide systemic upgrade of the power grid.

 

 

This is a typical example of "time frame mismatch":

Power plant construction can be completed relatively quickly, but grid upgrades are progressing slowly;

The demand curve continued to rise, but the pace of policies and projects failed to adjust in sync.

 

⸻

 

07｜The Correct Priorities in the Context of National Security

 

On the national security level, Cheng Maiyue agrees that electricity security can no longer be regarded as a simple energy issue.

However, he also cautioned that the priorities need to be further clarified in the wording—the issue is not about "giving it the same importance as other infrastructure projects," but rather that under pressure from AI and national security, the power grid should be "handled as a special case" and become a higher priority demonstration project.

 

The core logic is that when a key capability becomes a systemic constraint, delays will only amplify the risks and costs.

 

⸻

 

08 | (Opinion) The Rise of Electricity's Weight in the Context of Energy Transition

 

Later in the discussion, Cheng Maiyue also raised a more macro-level perspective:

As the energy transition drives the full electrification of end-use energy, the relative weight of electricity in the economic structure and national capacity is continuously rising.

 

He also emphasized that this transformation is essentially a dynamic process:

The prerequisite of "electricity consumption" must be met before it is possible to gradually replace high-carbon power generation with lower-carbon methods; if electrification is refused, the transformation goal itself will be impossible to achieve.

 

⸻

 

Only by thoroughly explaining the root cause can a solution be considered valid.

 

This pre-editing calibration further confirmed one point:

The so-called "power shortage in the United States" is not really about the amount of electricity generated, but rather the structural constraints on grid capacity and project time.

 

In the AI era, the key is not just "whether there is electricity".

The key lies in whether the power grid can meet the demand curve of the nation's capacity with the corresponding engineering speed and institutional efficiency.

 

⸻

 

Postscript |

 

This article is part of a GFM special study on "AI Energy," which aims to explore the structural challenges and feasible solutions facing the power system in the AI era from both engineering and institutional perspectives.

 

During the research and writing process, GFM also invited Mr. Cheng Maiyue, the column's host and consultant expert, to conduct pre-editing calibration and professional discussions on the relevant content, resulting in more detailed technical and engineering research materials.

 

Due to considerations of research ethics and usage scenarios, this part of the information has not been publicly released.

For academic, industry, or policy research needs, please contact us through GFM's official channels to learn how to use the relevant reference materials.