Artificial Intelligence,Cognitive Aging,and the Late-Career Surgeon: A Narrative Review

Cognitive Reserve

Cognitive domains central to surgical practice include processing speed, sustained attention, working memory, executive control, and visuospatial integration. These functions provide a framework for understanding how aging and artificial intelligence influence surgical judgment.

Cognitive aging is heterogeneous. Processing speed and working memory decline gradually, whereas crystallized knowledge and pattern recognition remain relatively stable.^17,18 Surgical decision-making—risk stratification, anticipation of complications, and intraoperative adaptation—depends heavily on experience accumulated over decades.

Cognitive reserve accounts for variability among surgeons. It describes the capacity to maintain performance despite age-related change. ¹⁸ Higher educational attainment and sustained cognitively demanding work are associated with delayed measurable impairment. ¹⁹ Surgical practice, which requires continual interpretation, anticipation, and correction, likely contributes to the development and maintenance of reserve across a career.

Reserve is strengthened through effortful engagement. Tasks requiring deliberate analysis and verification sustain executive function; passive acceptance does not.^20,21 The cognitive impact of artificial intelligence therefore depends on use: decision support systems that require active oversight may reinforce executive engagement and complement reserve, whereas automation that replaces analytic effort risks diminishing the cognitive activity that sustains surgical judgment.

Cognitive Offloading

One mechanism through which cognition may be influenced is cognitive offloading. Cognitive offloading refers to the use of external tools to manage information rather than relying solely on internal memory and recall. ²¹ Writing reminders, checklists, and calculators are familiar examples. These strategies improve efficiency and reduce mental strain. However, when information is routinely retrieved rather than actively reconstructed, independent recall and synthesis may weaken over time. ²⁰

Artificial intelligence represents an advanced form of cognitive offloading. Current systems include electronic health record–integrated documentation assistants, large language model drafting tools, automated chart summarization, and predictive analytics embedded in clinical workflows.^7–11 These applications generate documentation, synthesize longitudinal records, and distill complex data into structured summaries. Randomized trials, controlled implementation studies, and systematic reviews demonstrate reduced documentation time and clerical burden, lowering extraneous cognitive load.^12–16 These evaluations emphasize clinician verification and final judgment. The demonstrated benefit is workflow efficiency—not enhanced diagnostic reasoning or operative decision-making.

In surgical practice, substantial cognitive effort occurs outside the operating room. Documentation, electronic messaging, and regulatory requirements fragment attention and extend the workday. ²² By targeting these nonoperative demands, AI may preserve executive capacity for interpretation, anticipation, and intraoperative adaptation—the domains central to surgical judgment.

Whether AI strengthens or weakens professional reasoning depends on implementation. When surgeons actively review and revise AI-generated outputs, analytic engagement is maintained, consistent with mechanisms that support cognitive reserve.^18,19 When outputs are accepted without deliberate evaluation, reliance may increase and independent reasoning may attenuate.^20,21,23 In this respect, AI intersects with attention, working memory, and executive oversight rather than with technical skill.

Fatigue, Attention, and Performance

Fatigue influences surgical performance. Sleep deprivation slows reaction time, impairs coordination, and increases technical error. ²⁴ Extended work hours are associated with attentional failures, ²⁵ and procedures performed at night have been linked to higher complication rates. ²⁶ Distraction and interruption in the operating room further degrade performance. ²⁷

Cognitive strain accumulates over the course of the clinical day. Documentation demands, electronic messaging, and regulatory requirements add to operative stress and fragment attention. Late-career surgeons may retain sound judgment yet demonstrate reduced tolerance for sustained cognitive overload compared with earlier stages of practice.

AI tools that reduce clerical and organizational burden may therefore help preserve attention for tasks central to surgical care. Their value lies in limiting extraneous cognitive load and maintaining executive capacity—not in enhancing dexterity or altering technical execution.

Automation Bias

Surgeons routinely rely on decision support tools, including risk calculators, imaging reports, and laboratory reference ranges. These systems inform clinical reasoning. Concern arises when assistance becomes substitution. Automation bias describes the tendency to assign disproportionate weight to computer-generated recommendations over independent assessment.^23,28

As artificial intelligence becomes embedded within electronic health record workflows, its outputs may assume an appearance of authority or default acceptance. Under conditions of fatigue, time pressure, or cognitive overload, clinicians may be more likely to accept automated recommendations without deliberate scrutiny. ²⁸

AI intersects with surgical cognition at the level of attention and executive oversight. When outputs are critically reviewed, professional reasoning is maintained. When accepted without verification, independent assessment may attenuate. Whether AI strengthens or weakens surgical judgment depends less on the technology itself than on the discipline with which it is applied.

Physical Aging and Technical Limits

Beyond cognition, age-related changes in strength, fine motor control, reaction time, and visual accommodation are documented in both surgical and broader aging literature.^17,29–31 Reaction time and manual speed decline gradually with age, even among healthy professionals, and visual accommodation diminishes as part of normal physiology. Experience may compensate for some changes, but endurance and technical speed can be affected over time.

AI does not restore steadiness of hand, improve depth perception, or reverse slowing of motor responses. Operative performance remains dependent on physical capacity and technical skill. AI may reduce cognitive workload, but it does not alter the physiologic limits that shape technical execution.

Where AI May Support Late-Career Surgeons

AI is most likely to support late-career surgeons in information management rather than technical execution. Contemporary surgical practice requires continuous synthesis of imaging, laboratory trends, prior operative history, comorbid conditions, and evolving postoperative data. ²² Much of this work occurs outside the operating room and contributes to administrative and attentional burden. Systems that draft documentation, consolidate records, or summarize clinical information may reduce this extraneous load without supplanting clinical judgment.

Large language model–based tools generate structured documentation drafts for surgeon review and revision. Their demonstrated benefit is organizational efficiency. By streamlining nonoperative cognitive demands, these systems may preserve executive capacity for interpretation and intraoperative decision-making.

AI does not determine operative strategy, assess resectability, or execute technical maneuvers. Its contribution is cognitive rather than mechanical. When applied deliberately, it may support attention and consistency in reasoning. Its value lies in organizing information so that judgment can be applied more reliably—not in replacing that judgment.

Professional societies have recognized the limitations of self-assessment in late-career surgeons and recommend structured approaches to evaluating cognitive and psychomotor performance, particularly beginning in the seventh decade of practice.^32,33 These recommendations reflect consistent evidence that self-perception of decline does not reliably correlate with objective measures of performance, reinforcing the limitations of self-assessment in this population. ³²

Artificial intelligence offers a mechanism to operationalize this need. Consensus statements in surgical education support AI-based methods for objective assessment of technical performance and skill acquisition, provided that such systems are validated, explainable, and used to complement rather than replace expert judgment.^34,35 These tools extend naturally to longitudinal performance monitoring, where continuous data streams may allow earlier identification of change than episodic testing alone. ³⁴

At the same time, the role of AI in clinical decision remains supportive rather than substitutive, with current evidence demonstrating that machine learning models improve prediction.^36–38 This distinction is critical in the context of cognitive aging, where preservation of judgment depends on continued engagement in analytic processes rather than passive reliance on automated outputs.

The risks of automation bias must therefore be explicitly acknowledged. Studies of clinical decision support systems show that clinicians may overvalue algorithmic recommendations, particularly under conditions of fatigue or cognitive load.^23,28 In this context, AI may either reinforce or erode cognitive reserve depending on how it is used. Systems that require active verification and revision may sustain executive function, whereas uncritical acceptance of outputs risks attenuating independent reasoning.

Finally, the rationale for AI integration in late-career surgical practice is grounded in its ability to reduce extraneous cognitive burden. Time-motion studies demonstrate that a substantial proportion of physician effort is devoted to documentation and administrative tasks rather than direct clinical reasoning. ²² By offloading these demands, AI may preserve attentional capacity for interpretation, anticipation, and intraoperative decision-making—the domains most closely tied to surgical judgment.

Taken together, these findings suggest that AI should be incorporated into late-career surgical practice not as a means of compensating for cognitive decline, but as a structured adjunct to assessment and a tool for preserving cognitive capacity. Its value depends less on predictive capability than on the discipline with which it is applied.