AI brain fry: the allostatic load of working with AI

Everyone seems worried artificial intelligence is making us lazy, or stupid, or both. That we’ll all atrophy into passive content-consumers spoon-fed by the algorithm. I think that misses the real danger entirely. The problem isn’t that AI does the work for us; the problem is that it has fundamentally changed the kind of work we do. We’ve been shifted from the slow, deliberate work of creation to the frantic, relentless work of review. This constant, high-speed sifting and correcting is generating a novel form of cognitive exhaustion, a kind of AI brain fry that quietly racks up immense allostatic load on the nervous system.

Common Questions

What is 'AI brain fry'?

It's the specific cognitive exhaustion from working with AI. It’s not about being lazy; it's the fatigue that comes from being in perpetual high-alert review mode. You’re constantly making micro-decisions, correcting errors, and assessing outputs at a machine’s pace, which is a significant drain on your prefrontal cortex's resources.

Why is reviewing AI output so tiring?

Creating has natural pauses. You think, you write, you stare out the window. Reviewing a relentless stream of AI-generated content has none. Every output demands an immediate judgement: yes, no, tweak, discard. This constant decision-making without troughs of recovery is profoundly taxing on the architecture of your nervous system.

How does this relate to allostatic load?

Allostatic load is the cumulative wear and tear on your body from chronic stress (McEwen, 2017). Your body doesn't distinguish between being chased by a predator and being chased by 300 AI-generated emails needing your approval. The sustained vigilance and decision velocity feel the same, dysregulating your cortisol, metabolism, and immune function over time.

Related anchors: HRV anchor · burnt-out anchor · wired-tired anchor

The Bottleneck Has Moved, and It's You

The promise of AI was that it would free us from the drudgery of first drafts. And it has. But it’s replaced one form of labour with another that is far more metabolically expensive. The bottleneck in knowledge work is no longer production; it’s judgement. We used to spend our energy building something from scratch. Now, we spend it sorting signal from noise, correcting subtle but critical errors, and giving the final human sign-off on a torrent of machine-generated options.

This isn’t easier work. It’s faster work, and speed has a physiological price. According to recent research, the prefrontal cortex—the part of your brain responsible for executive functions like decision-making and impulse control—is a finite resource. Constant task-switching and evaluation deplete it, just as sprinting depletes your muscles. The difference is, your AI co-worker can sprint forever. You can’t. This is fundamental to understanding your body’s capacity for nervous system regulation.

Allostatic Load: The Real Invoice for Your 'Productivity Gains'

People talk about burnout as if it's a feeling. It isn't. Burnout is the bill arriving for months of unmanaged allostatic load. Coined by Bruce McEwen, allostatic load is the measurable, physical cost of chronic adaptation. When your body is constantly revving its stress-response systems—like the HPA axis (the brain-body circuit that releases cortisol)—without adequate recovery, things start to break.

Think of it as running a car's engine in the red for hours every day. Eventually, the gaskets blow. For humans, this looks like elevated blood pressure, insulin resistance, a weakened immune system, and a perpetually wired-but-tired feeling. As Rajita Sinha's work shows, this chronic stress isn't just in your head; it remodels your brain and body, making it harder to access rest and easier to slip into threat-detection (Sinha, 2008). Your body doesn’t know you’re just ‘reviewing a deck’; it only knows the pace, the vigilance, and the absence of a down-shift.

Speed is both a boon and a curse. The work is faster, but the human capacity for judgement isn't. Something has to give.

The Judgement Engine Needs Downtime: A Locus Coeruleus Story

Here's where we get nerdy. Your brain has a tiny region called the locus coeruleus (LC). Think of it as the gear-shifter for your brain's arousal and attention systems. When you need to focus on a task, the LC fires up, releasing noradrenaline to keep you alert and engaged. This is great for short bursts. But what happens when the task has no end?

Work by researchers like Mara Mather suggests that sustained, high-demand cognitive tasks can push the LC into a state of tonic, high-level firing (Mather, 2016). Instead of phasic bursts of attention followed by rest, it’s just stuck ‘on’. This is the neurobiological signature of hypervigilance. Your brain is locked in ‘scan for threats’ mode, even when the 'threat' is just a poorly phrased sentence in a generated report. This state is incredibly costly, burning through energy reserves and preventing the brain from entering the restorative states needed for memory consolidation and cellular clean-up. You're not just tired; your brain's basic maintenance routines are being cancelled.

Your HRV Is Your Judgement-Quality Score

If allostatic load is the long-term damage, Heart Rate Variability (HRV) is your daily dashboard indicator. HRV is the measure of the natural variation in time between your heartbeats, and it's a superb proxy for the state of your autonomic nervous system. High HRV means you’re adaptable, resilient, and in a state of 'rest-and-digest'—your body can fluidly shift gears. Low HRV means you’re stuck in a 'fight-or-flight' state, rigid and brittle.

Crucially, as Julian Thayer’s work demonstrates, there's a direct connection between HRV and the functioning of your prefrontal cortex (Thayer, 2009). When your HRV is low, your capacity for complex decision-making, emotional regulation, and clear judgement is impaired. You’re literally less intelligent. Trying to review complex AI outputs when your system is in a low-HRV state is like trying to edit a legal document after running a marathon. The errors you're supposed to catch are the very things your depleted brain will miss. Leaders who don't price this reality into their timelines are setting up their teams of burnt out employees for failure.

What to do this week

This isn't about ditching the tools. It's about architecting a human-centric way to work with them.

1. Bookend Your AI Sessions. Don't start your day with AI and don't end it with AI. Start with an analogue task (writing in a physical journal, planning your day) and end with one. Create a buffer between the machine's pace and your own.
2. Practice 'Review Sprints'. Batch your AI review work into focused 45-minute blocks. Then, take a mandatory 15-minute break where you get up, walk around, look out a window—anything but look at a screen. Use a timer. Don't cheat.
3. Introduce 'Dumb' Tasks. Schedule at least one task per day that requires zero intellectual judgement. Water the office plants. Tidy your desk. Walk to get a coffee. This gives your locus coeruleus and prefrontal cortex a chance to downshift. These aren't breaks from work; they are part of the work of recovery.
4. Embrace the Hard Stop. When the workday is over, it is over. The machine will keep offering more, but your biology requires a clean shutdown. The infinite scroll of work is a choice. Make a different one. For micro-resets, browse our library of 60-second hacks.

TL;DR

The narrative that AI makes us lazy is wrong; it's making us exhausted in a new way. By shifting knowledge work from creation to high-speed review, it places a relentless demand on our brain's finite capacity for judgement. This sustained vigilance drives up allostatic load—the cumulative physiological wear from chronic stress described by McEwen and Sapolsky—degrading our metabolic, immune, and cardiovascular health. The key is not to work harder, but to architect deliberate recovery troughs to buffer the machine's unrelenting pace.

Where this fits in the Kokorology system

This dynamic is a core contributor to systemic burnout and capacity collapse. Building resilience to it requires reinforcing the foundational pillars of physiological regulation. It's a key theme we address within our Anchor for Burnt Out Employees and the broader framework of Nervous System Regulation.

Closing

For leaders, the takeaway is stark. If you are calculating the productivity gains of AI for your workplace without also budgeting for the physiological recovery of the humans who must wield it, you are not seeing the full balance sheet. That productivity dividend will be paid back, with interest, in the form of attrition, sick days, and catastrophic burnout. Your team's capacity is not an infinite resource to be optimised; it is an ecosystem to be managed.

• Continue inside Performance L2, our advanced course on building sustainable capacity for leadership.
• Work with Sabin 1:1 to architect a bespoke capacity strategy inside private Coaching.
• Start with our Free Guide to Nervous System Regulation for the foundational principles.

Sources

• Mather, M., & Sutherland, M. R. (2016). The selective vulnerability of the locus coeruleus to hyperarousal and narcolepsy. Current Opinion in Neurobiology, 40, 151-158.
• McEwen, B. S. (2017). Neurobiological and systemic effects of chronic stress. Chronic Stress, 1.
• Sapolsky, R. M. (2004). Why Zebras Don't Get Ulcers: The Acclaimed Guide to Stress, Stress-Related Diseases, and Coping. Holt Paperbacks.
• Sinha, R. (2008). Chronic stress, drug use, and vulnerability to addiction. Annals of the New York Academy of Sciences, 1141, 105–130.
• Thayer, J. F., Hansen, A. L., Saus-Rose, E., & Johnsen, B. H. (2009). Heart rate variability, prefrontal neural function, and cognitive performance: the neurovisceral integration perspective on self-regulation, adaptation, and health. Annals of Behavioral Medicine, 37(2), 141-153.