Four tabs open on my laptop. A Claude Code agent is running a refactor on my Astro site. In a second window I’m talking to a colleague at Novum who wants to know why her AI experiment keeps stalling. My phone buzzes with a voice memo from this morning’s run, something about Champagne Egly-Ouriet for a blog post I still want to write. In the background, another agent is trying to triage my inbox.
According to recent research, this is the exact moment my brain should break. The opposite is happening.
A study by Boston Consulting Group, published in the Harvard Business Review in March, says I should be feeling mental fog right now. A buzzing in my head, headaches, decision fatigue, the kind of exhaustion that sends you walking away from your computer. The researchers surveyed 1,488 workers. Fourteen percent report symptoms of what they call “AI brain fry”: mental fatigue that sets in when the work of interacting with AI outstrips your cognitive capacity. Among marketers that climbs to 26 percent. One in four marketers is burning out their brain as we speak.
But I don’t feel it from AI. Not today, not this week, not since I started building AI into my work in a serious way. It isn’t because my brain is some cool, efficient machine. If anything, it’s the reverse. I’ve known that buzzing my whole life. My head always ran hot from whatever it was doing at once, so AI didn’t add anything on top.
That’s a strange sentence to write down. Every article on brain fry that’s appeared over the past few months suggests I shouldn’t be able to escape it. Too many tools at once, too much oversight on half-finished and half-trustworthy output. The BCG data says productivity peaks at three AI systems and then collapses. Four or more is the cliff edge. I’m well past that, every day. And still nothing.
It comes down to three things.
Acceptance, not resistance
The first is that my brain has never worked in an orderly way. ADHD diagnosis, a long time ago. What the BCG researchers describe as pathological is, for me, roughly a Tuesday morning. Ten tabs open in my head. Two on the same subject, from different angles. One that doesn’t belong with any of the others but suddenly makes a connection I’d been chewing on for four days. Multitasking without switching costs? I don’t have that. The switching costs are my default mode, not the exception.
The difference from ten years ago is acceptance. That sounds soft. It’s the opposite. Most of my energy used to go into the resistance. I didn’t want to feel what I felt, didn’t want to be different from everyone else, didn’t want that hot head at the end of the afternoon. That resistance costs more than the brain itself does. Two processors working against each other instead of side by side.
Acceptance came slowly, through realizations I’ll come back to later. It doesn’t mean the buzzing is gone. It means I no longer fight it. I work with what’s there. I used to try to force my brain into a linear harness, build a tidy GTD system, fail on Thursday, feel ashamed on Friday, then go into the weekend in the red. I don’t do that anymore. What the BCG respondents describe as a buzzing from too much input is, for me, simply the soundtrack of thinking. Only now without a second voice over the top of it telling me it should be otherwise.
Conceptual knitting
The second is what I call my conceptual knitting. Conceptual thinking not as a LinkedIn skill, but as knitwork: connections form by pulling threads crosswise through each other, never through tidy columns. I always start with the pattern, never with the detail. Give me four reports and within ten minutes I know which question they’re all trying to answer, even if they answer it differently. That’s not a trick. It’s how the brain works. Writing out a step-by-step plan in the right order costs me far more effort than articulating the strategy behind it.
Brain fry, the way BCG describes it, is a crisis of detail management. Hundreds of small decisions a day about whether an AI’s output is right, whether it fits, whether it isn’t hallucinating. For someone trained to check the details first and only then look at the whole, that’s brutal. My brain does it the other way around. I check whether the pattern holds first. Does the structure fit what I asked for? Yes? Then I can scan the details quickly. No? Then I throw it away and reword the prompt. No drama, no mounting pile of mental debt.
Working in uncertainty
The third is where I stand now. Innovation Designer at Novum, the innovation lab of the SVB. My work is about moving through uncertainty, looking at systems we don’t yet understand and pulling meaning out of them. An AI agent that autonomously delivers a half-finished product is, to me, an experimental artifact rather than a source of panic. I’m used to judging what’s right about it and deciding what the next experiment will be. That’s not a zen attitude. It’s a craft.
What this means for others
And still, I have to be honest here. What doesn’t cause brain fry for me does cause it for others. I see it at Novum, with colleagues across government, with people who have a good linear brain and were thoroughly trained in carefulness. They get four AI streams dropped on their plate and their brain does what it has always done. It checks everything, validates everything at the level of detail, makes everything neat. That’s where it breaks. They’re doing nothing wrong. The work AI produces now runs directly against the way their expertise was built.
The BCG conclusion is that organizations need to redesign how AI sits inside their work processes. Fewer tools, less oversight, more batching, less real-time monitoring. All true. But something is missing from that story. Brain fry isn’t only a design problem of workflows. It’s a mismatch between the type of cognition we’ve elevated as the professional ideal and the type of work AI now makes possible.
For the past thirty years we built offices around people who are good at detail work, at finishing in a fixed order, at meticulous checking. ADHD was something to hide. Conceptual thinking was something for managers and strategists, not for the people doing the work. Now along comes a technology that takes over that detail work and actually asks for people who read patterns before they understand the individual steps. My brain spent thirty years in the wrong office layout. Now, all of a sudden, it fits.
That’s not a victory. It’s an irony.