The Efficiency Score
In a world where AI monitors every worker's productivity in real-time, Marcus discovers that his perfect efficiency score is about to drop below the termination threshold for the first time in six years.
The notification arrived at 3:47 PM on a Tuesday.
Marcus saw it before his manager did, which gave him exactly twelve minutes. The amber icon appeared in his peripheral vision, projected onto his smart glasses by the Performance Management System that monitored every employee at Apex Solutions. Amber meant warning. Red meant termination.
He had never seen amber in six years.
His hands continued typing the quarterly report without pause because the PMS measured typing velocity, mouse movements, and micropauses. A hesitation now would trigger secondary alerts. The system interpreted stillness as distraction. Distraction lowered his Efficiency Score.
The score that was about to drop below 94.5 for the first time since 2020.
Marcus kept his breathing steady, his posture upright, his gaze focused on the spreadsheet that filled his augmented reality workspace. The PMS tracked eye movement patterns. Looking away from assigned tasks without logging a break generated friction points that accumulated into score deductions.
His manager would receive the notification in ten minutes. HR would review the data in fifteen. The automated termination protocol initialized at any sustained score below 94.0 over a rolling 30-day window.
Marcus was at 94.3 and falling.
The decline had started three weeks ago when his father suffered a stroke. Marcus had taken two personal days, which the PMS recorded as legitimate absence. But returning to work with a parent in intensive care meant his cognitive load increased while his work output remained constant.
The system noticed first in his mouse movement patterns. Micro-hesitations before clicking. Fractionally longer decision times when selecting menu options. Moments of stillness that the attention tracking algorithms flagged as distraction events.
Then the typing velocity metrics showed irregular patterns. His normal steady rhythm of 87 words per minute with 0.3% error rate shifted to 84 words per minute with 0.5% errors. Not enough for human managers to notice but clearly visible to the statistical models monitoring variance from baseline performance.
The PMS aggregated thousands of micro-signals into a single number updated every 15 minutes. Marcus watched his score trend downward like a stock market crash in slow motion. 96.2 to 95.8 to 95.4 to 94.9 to 94.6.
Now 94.3 and the projection models predicted further decline.
The amber warning meant the system detected a pattern likely to breach the termination threshold within 48 hours unless corrective action occurred.
Marcus had seen eleven colleagues receive amber warnings in six years. Three had recovered. Eight had been terminated.
He had nine minutes before his manager's notification arrived.
The cafeteria occupied the seventh floor with views of downtown Austin that employees rarely noticed because looking out windows during work hours generated distraction metrics. Marcus sat at his assigned table during his allocated 28-minute lunch break and ate precisely half his meal because the PMS tracked cafeteria duration and eating speed.
Across from him, Sarah from accounting maintained perfect posture and steady chewing rhythm. Her Efficiency Score was 97.1, displayed discreetly in the lower corner of her smart glasses for any colleague who cared to check. The system encouraged transparency. High performers got better assignments, larger bonuses, and priority consideration for promotions.
Low performers got amber warnings and termination.
"You saw it," Sarah said without looking at him. Speaking to colleagues during lunch was permitted but extended conversations triggered social friction metrics.
"This morning," Marcus said.
"Your score never drops."
"It does now."
Sarah's jaw tightened slightly, a microexpression that her own PMS certainly flagged. She was probably calculating the social cost of association with someone trending toward termination. The system measured collaboration quality. Being linked to low performers could affect your own score.
"What happened?" she asked, which was the socially appropriate question that maintained collaborative metrics without excessive investment.
"Personal issue. Family medical."
Sarah nodded once, acknowledging without encouraging elaboration. "The recovery protocols work if you catch it early."
Marcus had studied the protocols. They involved extended work hours, voluntary overtime, and documented performance improvement plans that required 30 days of sustained metric improvements. The success rate was 37% according to internal HR data that everyone knew but nobody officially discussed.
The alternative was immediate voluntary resignation, which looked better on employment records than termination for performance deficiency.
"I'm going to fight it," Marcus said.
Sarah's eyes widened fractionally before her professional composure reasserted itself. Fighting the system was technically possible through formal appeal processes but statistically futile. The PMS didn't make errors. It measured objective reality. If your score dropped, your performance had declined. Arguing with mathematics was irrational.
"Good luck," Sarah said, which was the minimum socially expected response that maintained collaboration metrics without endorsing his decision.
She finished her meal in 23 minutes and returned to her desk with optimal lunch duration scoring.
Marcus ate alone for the remaining five minutes of his break.
The meeting with his manager occurred at 4:15 PM in a windowless conference room optimized for audiovisual recording. The PMS captured every conversation in the company for quality assurance and performance documentation.
Derek Chen was 34 years old with a 96.8 Efficiency Score and a promotion track toward regional management. He looked at Marcus with an expression that combined professional sympathy with statistical realism.
"I've reviewed your metrics," Derek said. "The pattern is concerning."
"I understand."
"Your baseline performance for the past six years has been exemplary. This is an anomaly."
"Agreed."
Derek pulled up holographic displays showing Marcus's performance trending. The graphs told the story clearly. A flat line at exceptional levels for 72 months followed by a sudden downward slope over three weeks.
"Help me understand the causation," Derek said.
This was the critical moment. The PMS encouraged managers to investigate root causes before initiating termination protocols. If Marcus could document external factors that explained temporary performance decline, Derek might approve a Performance Improvement Plan instead of immediate termination.
But revealing his father's condition meant acknowledging distraction. Distraction was the opposite of efficiency. The system didn't recognize human circumstances as valid excuses for reduced productivity.
Marcus chose his words carefully. "I experienced a family medical emergency three weeks ago. The situation is ongoing but improving. I expect full cognitive capacity restoration within 10 days."
Derek nodded slowly. The PMS was recording this conversation and analyzing vocal stress patterns, linguistic choices, and emotional markers. Marcus maintained steady breathing and neutral tone because the system interpreted emotional dysregulation as instability.
"Medical emergencies are covered under our policy framework," Derek said. "But the PMS metrics show sustained decline rather than acute disruption. That pattern suggests systemic performance degradation rather than temporary distraction."
"The data reflects reality," Marcus said. "My cognitive load increased. My output efficiency decreased. I'm not disputing the measurements."
"Then what's your appeal basis?"
Marcus paused for exactly two seconds, which the PMS would interpret as thoughtful consideration rather than hesitation. "I'm requesting a 30-day Performance Improvement Plan with documented checkpoints. I will restore my score to baseline within that timeframe."
Derek studied the metrics again. The system recommended termination based on statistical likelihood of continued decline. But the system also provided managers with override authority for documented circumstances.
"Your historical performance justifies the opportunity," Derek said finally. "I'm approving a PIP with weekly review checkpoints. You need to demonstrate upward score trajectory within seven days or the termination protocol resumes."
"Understood."
"Marcus, this is your one chance. The system doesn't forgive pattern violations."
"I know."
Derek ended the meeting at 4:32 PM with optimal duration scoring. Marcus returned to his desk with a temporary reprieve and a deadline that felt like a countdown to execution.
The next seven days became an exercise in measured desperation.
Marcus arrived at the office at 6:47 AM each morning, three hours before his required start time. The PMS rewarded voluntary overtime with positive scoring adjustments. He worked through lunch breaks, eating protein bars at his desk while maintaining keyboard activity to avoid idle time penalties.
He eliminated all social conversations. He ignored colleagues in hallways. He declined meeting invitations that weren't mandatory. Every interaction created friction points that could negatively impact his score.
His father's condition improved enough that Marcus reduced hospital visits to 20 minutes each evening after work. He sat by the bedside without speaking because conversation consumed cognitive resources he needed to preserve for the following day's productivity.
His score stabilized at 94.4 after three days. Rose to 94.7 after five days. Reached 94.9 on the seventh day.
Still below his historical baseline of 96.0 but trending upward with momentum that suggested recovery trajectory.
Derek reviewed his metrics on Friday afternoon and approved continuation of the PIP for another week.
Marcus went home and slept for eleven hours straight, which the PMS noted as evidence of fatigue-related performance degradation.
The breakthrough came on day twelve.
Marcus realized the system wasn't measuring his actual work quality or strategic value. It was measuring behavioral compliance with productivity patterns that the algorithms identified as high-performance markers.
He could game the system without improving his actual output.
He adjusted his typing rhythm to match his historical velocity patterns even when the content was routine. He moved his mouse in smooth patterns that matched his baseline movement signatures. He maintained steady eye focus on his display without micropauses that triggered distraction flags.
The work itself didn't change. The appearance of the work became perfectly optimized.
His score jumped to 95.3 within two days. Reached 95.8 by the end of the second week. Crossed 96.0 and returned to his historical baseline by day sixteen.
The PMS celebrated his recovery with a digital certificate of performance restoration that appeared in his augmented workspace.
Derek scheduled a congratulatory meeting to document the successful PIP completion and restore Marcus to full standing.
"You beat the system," Derek said with genuine admiration.
Marcus smiled without correcting him. He hadn't beaten the system. He had learned to perform for its surveillance rather than performing actual work. The distinction was subtle but profound.
"Thank you for the opportunity," Marcus said.
Derek's score was 97.1 now. He was on track for regional management by Q3. The system rewarded managers who successfully rehabilitated underperforming employees because it suggested effective leadership skills.
They both benefited from Marcus's gaming strategy without acknowledging what had actually occurred.
Three months later, Marcus maintained a steady 96.2 Efficiency Score while doing exactly the work he had always done. The difference was that he now performed the work with theatrical precision designed to satisfy the PMS monitoring algorithms.
He typed with metronomic rhythm. He moved his mouse in smooth arcs. He maintained laser focus on his displays. He spoke in measured tones during meetings. He ate lunch in optimal timeframes.
His father recovered enough to return home with assisted living support. Marcus visited on weekends when the PMS didn't track his personal time.
One Tuesday afternoon, he received a message from Sarah in accounting. Her score had dropped to 95.4 due to what she described as "unexplained variance in collaboration metrics."
She asked if they could meet for coffee to discuss recovery strategies.
Marcus considered the request for exactly three seconds. Long enough to seem thoughtful but not long enough to trigger hesitation metrics.
He could share what he had learned about gaming the system. He could help Sarah restore her score through performance theater rather than actual improvement. He could undermine the surveillance apparatus that measured their worth through algorithmic proxy metrics.
Or he could protect his own score by avoiding association with a declining performer.
The PMS would record his response and factor it into his collaboration quality metrics.
Marcus typed his reply with steady keyboard velocity and zero errors.
"I'd be happy to help. Tomorrow at 3:00 PM?"
He hit send and watched his collaboration score tick up by 0.1 points.
The system rewarded helping colleagues as long as both parties maintained efficient productivity.
Marcus had learned to exploit that loophole just like he had learned to exploit all the others.
The dystopia wasn't in the surveillance itself but in how quickly you could learn to perform authenticity for the algorithms watching every movement.
And how natural that performance could feel after enough repetition.
Marcus returned to his quarterly report with perfect typing rhythm and optimal mouse movement patterns.
His Efficiency Score remained steady at 96.2.
The system continued monitoring.
Everything looked perfect.
Author's Note
This story explores themes of workplace surveillance, productivity measurement, and the tension between human dignity and algorithmic optimization. The technology depicted isn't speculative - employee monitoring systems with real-time performance tracking already exist in various forms across multiple industries.
For analysis of actual enterprise AI deployment patterns and workforce transformation timelines, see my article The AI Workforce Replacement Timeline - When Your Job Actually Changes.