The technical subject chosen for this stress test is characteristic mode analysis, or CMA, a powerful mathematical framework used to decompose the currents and fields on an electromagnetic structure into a set of orthogonal modes with well-defined energetic properties. Originally developed for conducting bodies in the early 1970s, CMA has evolved into a practical design methodology for antennas and, more recently, for metasurfaces, which are engineered arrays of subwavelength elements on electrically thin surfaces. These surfaces can steer beams, focus radiation, convert polarization, absorb energy, and filter spatial frequencies across regimes spanning microwaves to optics. By tailoring the local boundary response of a surface, engineers can implement functions that would be difficult or impossible with conventional materials.

The core appeal of CMA lies in its interpretability. When a perfect electric conductor is discretized using the method of moments, the resulting impedance matrix can be split into real and imaginary parts representing radiation and reactive energy storage, respectively. Solving a generalized eigenvalue problem then yields a set of characteristic currents, each associated with an eigenvalue whose sign and magnitude reveal whether that current pattern is predominantly storing or radiating energy. An eigenvalue near zero indicates self-resonance, positive values suggest inductive behavior, and negative values indicate capacitive tendencies. From these eigenvalues, engineers derive practical metrics such as modal significance, which ranges from zero to one and peaks near unity when a mode is strongly radiating, and the characteristic angle, a phase-like quantity that crosses 180 degrees at resonance under common conventions.

What makes CMA particularly valuable for metasurface design is the conceptual separation it provides between what a structure can support and what a given excitation actually drives. The induced current on a metasurface can be approximated by a truncated modal expansion whose coefficients depend on the feed, port, or incident field. This means that modes with high modal significance are more readily excited, and feed placement or polarization determines which modes contribute to the final response. For metasurface antennas, this separation supports systematic feed design: a feed that overlaps strongly with the desired mode’s current distribution while minimally coupling to parasitic modes tends to yield cleaner radiation and more predictable tuning behavior. For illuminated metasurfaces such as reflectarrays, transmitarrays, and reconfigurable intelligent surfaces, the incident polarization and angle act as the excitation, allowing CMA to diagnose which modes couple to a given illumination and which symmetries must be broken to enable the target response.

The review organizes CMA applications in metasurface design into four recurring roles. First, CMA diagnoses dominant physics by identifying which currents and resonances actually matter in a given structure. Second, it shapes bandwidth and radiation patterns through multi-mode control, where broadband performance can benefit from multiple modes with high modal significance across adjacent frequency regions. Third, it enables reconfigurability by tracking how modal trajectories shift across switching states in devices such as reconfigurable intelligent surfaces. Fourth, it reduces design iteration by providing interpretable guidance on feed placement and geometric perturbation, replacing brute-force optimization with physics-informed reasoning. Geometric modifications such as slotting, truncation, adding vias, changing periodicity, or introducing resistive and reactive loading can move modal resonances, split degenerate modes, or alter current distributions, thereby changing the achievable scattering and radiation response.

Concrete examples from the literature illustrate these principles in action. Researchers have used CMA to suppress higher-order modes in closely spaced phased arrays through metasurface placement, to design dual-band metasurface antennas by ensuring that two suitable modes resonate in distinct frequency bands, and to develop Ka-band metasurface antenna arrays with controlled sidelobe levels and improved aperture efficiency. In polarization conversion, CMA identifies modal current symmetries and their relative phases under excitation, explaining when cross-polarized reflected fields arise under plane-wave illumination. For frequency-selective surfaces, periodic CMA reveals which resonances control the spectral response and how resistive loading shifts and damps those resonances, as demonstrated in experimentally verified studies of resistor-loaded structures. Notably, recent advances have shown that characteristic modes of periodic systems can be extracted directly from scattering parameter data, making CMA accessible from generic solvers or even measurements.

The review also addresses a recurring pitfall in metasurface engineering: the assumption that unit-cell analysis under periodic boundary conditions fully predicts the behavior of a finite aperture. Periodic CMA is valuable for identifying local resonances and designing target reflection or transmission phase responses, but finite apertures introduce edge diffraction, non-uniform excitation, mutual coupling variations, and truncated periodicity effects that can shift modal content and bandwidth. The authors emphasize that periodic insights should be validated by finite-aperture analysis or sufficiently large supercell simulations, checking whether the same dominant modes persist under realistic excitation. Similarly, in lossy structures involving conductor loss, dielectric loss tangent, or lumped element dissipation, the classic orthogonality properties of characteristic modes can break down and eigenvalues may become complex, altering the meaning of derived metrics. These are not failures of CMA itself but reminders that the assumptions behind a metric must match the physics of the metasurface being analyzed.

Beyond conventional design, the review highlights how CMA connects to cutting-edge approaches involving global optimization and machine learning. For multilayer or volumetric metamaterials, useful behavior may arise from strong near-field coupling among electric and magnetic resonators distributed across multiple layers or throughout a three-dimensional lattice. In such cases, a purely local or two-dimensional periodic modal picture may miss global current pathways, interlayer hybridization, and edge-mediated coupling. The authors point to recent work on superradiant broadband magnetoelectric arrays empowered by meta-learning and on evolutionarily designed three-dimensional metamaterials for scattering maximization, where optimization methods such as covariance-matrix adaptation evolution strategies discover strongly coupled configurations whose electric and magnetic current families overlap spectrally and radiate with compatible phases. This principle, termed resonance cascading, deliberately arranges several nearby resonances so that they interfere constructively over the band of interest, sustaining enhanced scattering over a broader bandwidth than any single isolated resonance could achieve.

The methodological audit component of the paper is perhaps its most consequential contribution. The authors identify three highest-risk failure modes in AI-assisted technical writing: hallucinated or incorrect references, correct references but incorrect attribution of claims, and subtle technical inconsistencies such as conflating periodic unit-cell results with finite-aperture behavior without qualification. To address these risks, the manuscript includes an explicit audit that verifies each reference entry against external metadata and checks key equations, symbol definitions, and cross-section consistency. The authors recommend that AI-assisted reviews adopt a similar trust-but-verify protocol, documenting what was verified and what remains interpretive. They stress that this should not be interpreted as evidence that AI replaces domain experts; rather, it motivates a workflow in which AI accelerates early drafting while domain experts retain responsibility for technical correctness, scope control, and claim-to-citation alignment.

Looking forward, the authors anticipate progress in generalized CMA formulations for lossy, dispersive, and strongly coupled periodic structures, along with community benchmarking and reporting practices that make CMA-guided metasurface designs more reproducible across solvers and laboratories. They envision combining CMA-based physical interpretability with global optimization tools, particularly for multilayer or volumetric scatterers where constructive interference among several resonances can be designed rather than guessed. The paper’s figures are explicitly labeled as illustrative schematics and synthetic plots used to explain modal reasoning rather than to report new electromagnetic simulations or measurements, maintaining transparency about the boundary between tutorial synthesis and original research. As AI tools become increasingly embedded in scientific workflows, this study offers a concrete template for maintaining scientific reliability: embrace the speed and fluency of automated drafting, but anchor every claim to verified evidence and every equation to checked physics.

Subject of Research: AI-assisted technical writing and characteristic mode analysis for metasurface design

Article Title: Can we trust AI-assisted technical writing in electromagnetics? characteristic mode analysis for metasurface design as a stress test

Article References: Hamdalla, M. Z. M., & Roy, S. (2026). Can we trust AI-assisted technical writing in electromagnetics? characteristic mode analysis for metasurface design as a stress test. Frontiers in Antennas and Propagation, 4, Article 1808402. https://doi.org/10.3389/fanpr.2026.1808402

Image Credits: AI Generated

DOI: 10.3389/fanpr.2026.1808402

Keywords: characteristic mode analysis, metasurfaces, AI-assisted writing, antenna engineering, reconfigurable intelligent surfaces, frequency-selective surfaces, polarization conversion, eigenmode decomposition, reproducibility, citation verification, electromagnetics, global optimization

The inspiration for the framework came, in part, from a moment of professional dismay. At an academic conference, Giray watched a business professor present what was framed as a breakthrough in AI-assisted pedagogy: students would receive a case study, paste it into a generative AI tool, lightly edit the output, and submit the result. The audience applauded. For Giray, the episode crystallized a troubling trend he describes as an AI-first, think-second culture, in which students ask before they think, researchers prompt before they hypothesize, and writers generate before they reflect. In such a design, writing, long the traditional proxy for human understanding, is reduced to post-processing, and the friction of articulation through which ideas are actually formed is removed entirely.

The SSS framework is anchored in three bodies of scholarship. The first is philosophical: Marcus Aurelius’s distinction between what lies within our control and what does not. Applied to AI, the principle directs attention inward, toward the deliberate exercise of cognitive agency, rather than outward, toward futile resistance to the technology’s advance. The second foundation is empirical cognitive science, specifically Robert Bjork’s concept of desirable difficulties, which holds that learning conditions that feel effortful in the moment produce more durable and transferable knowledge than fluent, frictionless ones. When learners generate their own answers and wrestle with problems, they encode knowledge more deeply than when information is delivered efficiently and passively. The third pillar is the extensive literature on automation bias, the well-documented tendency of users to defer to algorithmic outputs even when the machine is demonstrably wrong.

The framework’s name is a deliberate metaphor borrowed from competitive gaming, where an SSS rank denotes the highest achievable performance level. The model is designed to optimize conditions for producing the best possible output through three phases: Set an AI-Free Zone, Struggle, and Spar with AI. Each phase is grounded in specific empirical findings, and each carries explicit boundary conditions acknowledging that the depth of engagement should scale with the complexity and novelty of the task at hand.

The first phase is environmental and behavioral. Because willpower is a depletable resource, and because the mere presence of tempting stimuli reduces cognitive capacity, the framework prescribes situation modification rather than resistance: physically separating AI-accessible devices from the workspace, re-centering analog tools such as paper and pen, and reconfiguring the environment for slow, uninterrupted reasoning rather than speed. Support for this phase comes from a striking study by Nataliya Kosmyna and colleagues, in which 54 participants drafted essays under different conditions. Those writing unassisted showed the strongest neural connectivity, superior memory recall, and the highest sense of ownership over their text. Critically, participants who transitioned from unassisted writing to AI assistance preserved those cognitive advantages, while those who started with AI showed attenuated brain connectivity, suppressed attentional networks, and reduced content retention.

The second phase, Struggle, is the cognitive core of the framework: sustained, unassisted ideation, drafting, and sense-making before any AI consultation. The term is chosen deliberately. Desirable difficulty research demonstrates that the effort of retrieval and generation, even when it produces incomplete and messy output, creates stronger encoding than passive reception of polished information. The framework embraces Anne Lamott’s celebrated notion of the imperfect first draft, treating incompletion not as failure but as the necessary precondition for genuine intellectual ownership. Practical sub-strategies include translanguaging, allowing thought to flow across languages and registers; non-linear ideation through scribbling and visual mapping; and tolerance of incompletion, remaining with a problem long enough for non-obvious connections to emerge.

The most compelling empirical validation comes from a controlled study by Wong and Qiu involving 196 university students who were assigned to one of three AI collaboration protocols: human-only, general free use of ChatGPT, or a regulated think-first structure. On the immediate creative task, the unrestricted AI group outperformed the others. But on a subsequent independent task completed without any AI assistance, the regulated group significantly outperformed both alternatives in independent originality. Process analysis revealed the mechanism: the general-AI group primarily instructed the AI to generate solutions directly, while the regulated group used AI to improve and challenge ideas they had already generated themselves. The conclusion is that generating one’s own ideas first, then using AI to refine them, yields durable learning gains that passive AI use does not.

The third phase, Spar with AI, borrows its logic from martial arts, where a sparring partner does not fight for the practitioner but tests them, exposing weaknesses and demanding articulation of positions under pressure. Here AI is engaged as an active intellectual interlocutor rather than a content generator. Productive uses include asking the AI to identify the strongest counterarguments to a human-drafted argument, detect logical gaps, refine terminology, review structure, and expand perspectives. What the phase explicitly excludes is delegation of thinking: AI does not write the argument, generate the thesis, or replace the writer’s voice. When that relationship inverts, the anchor effect reasserts itself, an early AI output colonizes subsequent human reasoning, and automation bias resumes its corrosive work.

The editorial also situates the framework within a set of ethical concerns that extend beyond individual cognition. At the collective level, widespread reliance on AI-generated content risks what critics have termed AI slop, minimally human content that produces an epistemic monoculture in which models trained on model output amplify existing biases and narrow the range of circulating ideas. AI language models additionally encode documented patterns of linguistic discrimination against non-native English writers and cultural bias toward Western rhetorical norms, meaning practitioners who surrender initial drafting to AI risk having their distinctive voices smoothed away. Neil Postman’s warning about the deification of technology, the cultural disposition to treat machine outputs as inherently superior to human judgment, looms over the entire discussion, and the framework is presented as a structural inoculation against it.

Giray is candid about the framework’s limitations. The three-phase structure has not been directly validated as a unified model through experimental research, even though each component draws on substantial empirical support from desirable difficulty, situation modification, cognitive forcing functions, and designed friction in human-computer interaction. He proposes a phased research agenda beginning with survey studies to map practitioner behavior, followed by randomized controlled trials comparing SSS-aligned workflows with AI-first alternatives, mixed-methods analyses of how outcomes vary with task complexity and expertise, and long-term cohort studies, including multilingual and non-Western populations. A faculty meeting toolkit accompanies the paper to help educators translate the sequence into concrete assignments. The deeper proposition, however, stands on its own: the human brain operates on roughly twenty watts of metabolic energy yet achieves associative reasoning and ethical judgment that no algorithm replicates, and those capacities require exercise. Clear the space, wrestle with the idea, and only then invite AI to challenge what human thinking has produced. In that sequence, AI may magnify the reach of human thought, but the power to judge must remain with the thinker.

Subject of Research: A three-phase framework for sequencing human cognition before AI assistance in writing

Article Title: SSS Framework: A Think-First, AI-Second Model for AI-Assisted Writing

Article References: Giray, L. (2026). SSS Framework: A Think-First, AI-Second Model for AI-Assisted Writing. Annals of Biomedical Engineering. https://doi.org/10.1007/s10439-026-04358-5

Image Credits: AI Generated

DOI: 10.1007/s10439-026-04358-5

Keywords: SSS framework, AI-assisted writing, human-AI collaboration, automation bias, desirable difficulties, cognitive offloading, intellectual autonomy, productive friction, generative AI, cognitive science, education, epistemic erosion