The Second Neural Network — Why Claude Keeps You Awake at Night
Last week I woke up at 3am. My first thought: I still need to send a prompt. That's not productivity — that's a signal. One that every leader needs to understand.
Tom Adebahr
Thought Leadership KI
- KI baut in deinem Kopf ein zweites kognitives Netz auf — schnell, unsichtbar, und mächtiger als du denkst.
- Wie im Sport gilt: Wachstum passiert in der Erholung, nicht im Training. Wer ohne Rhythmus promptet, wird dysreguliert.
- Kognitive Dysregulation durch KI-Übernutzung ist real — und hat konkrete Symptome, die Leader erkennen müssen.
- AI builds a second cognitive network in your mind — fast, invisible, and more powerful than you realize.
- As in sport: growth happens in recovery, not in training. Prompting without rhythm leads to dysregulation.
- Cognitive dysregulation from AI overuse is real — and has concrete symptoms that leaders must recognize.
- Leadership 2026 means: consciously managing your own AI usage and your team's — not less AI, but with rhythm.
When you work intensively and regularly with AI models like Claude, it changes how you frame problems, how you think and how you make decisions — even when no AI model is open. The second neural network is the cognitive layer this collaboration builds inside you: a kind of mental infrastructure for AI-assisted thinking. It makes you faster, clearer and more effective. But like any network, it needs maintenance, boundaries and recovery time — otherwise it overgrows the slow, creative, intuitive thinking that is at the core of leadership.
How AI Builds a Second Brain Inside You
It starts harmlessly. First prompts, first surprises. You see what Claude can do and start approaching tasks differently. You frame problems to be workable for AI. You think in delegation rather than execution. You run thought processes as if you were prompting them — even when sitting alone.
This isn't a side effect — it's the mechanism. Neuroplasticity working for us. Those who interact daily with a high-performance thinking system internalize its logic. The brain learns to work more efficiently because it knows a system that can fill gaps. You get faster. Clearer. Capacity freed up for strategic thinking.
But there's a flip side. Those who delegate every piece of research to Claude lose tolerance for the slow searching that sometimes produces the best ideas. Those who prompt every text lose practice in the laborious struggle for the right words — a struggle that creates depth. Those who outsource every analysis lose the ability to sit with uncertainty without immediately wanting an external answer.
The signal for me was the 3am moment. I had woken up and my first thought wasn't about a problem I needed to solve — it was about a prompt I hadn't sent yet. That's not a sign of productivity. That's a sign that the second neural network had started to take over.
"AI connects with your thinking. Faster than you notice. That's the advantage — and the risk at the same time."
Tom AdebahrThe Athlete Analogy: Stimulus + Recovery = Growth
Anyone who takes training seriously knows the principle of supercompensation: the body doesn't grow during training. It grows in the recovery afterwards. Training sets the stimulus — recovery is the response in which adaptation takes place. Those who ignore this principle and train every day at Hyrox intensity don't end up in peak form. They end up overtrained.
Overtraining in sport is well documented: performance decline despite training volume, sleep disturbances, irritability, loss of motivation, increased injury susceptibility. In extreme cases, rhabdomyolysis occurs — the breakdown of muscle tissue that damages the kidneys. The analogy for cognitive overloading through AI is not academic. The symptoms are the same.
Periodization: What Elite Athletes Know Better Than Leaders
Elite athletes work with periodization plans. They alternate high-intensity phases with active recovery, deload weeks and targeted regeneration. Not because they're weak — but because they understand that performance over the long term comes from intelligent management, not maximum load at every moment.
No world-class runner trains at race pace every day. No weightlifter lifts their maximum every day. The best results come from the structured alternation of tension and release. That's not a soft skill — that's physiology.
For AI usage this means: 16 hours of prompting, optimizing, delegating — without rhythm — doesn't make you faster. It makes you dysregulated. The curve rises steeply. Then it collapses. And the collapse is the deeper the steeper the rise was.
Dysregulation — The Silent Risk of Intensive AI Use
Cognitive dysregulation from AI overuse is harder to recognize than physical overtraining. There's no muscle soreness, no obvious injury picture. The symptoms are subtler — and therefore more dangerous.
Symptoms Leaders Need to Recognize
Decision fatigue despite AI support: You have more options than ever before — Claude generates alternatives in seconds — but you find it harder to decide than before. The number of possible paths paralyzes rather than liberates. This is a classic sign of cognitive overloading.
Always-on mentality: You no longer switch off. In the evening at a restaurant you're thinking about how you could prompt the table conversation. During sport your head runs through workflows. At night thoughts send you to your phone. The second neural network runs in the background — permanently, even when you're resting.
Loss of slow thinking: The ability to sit with a question for a long time without immediately wanting an answer erodes. You become impatient with processes that deliver no immediate outputs — deep conversations, books, quiet reflection. But these are precisely the processes from which strategic leadership emerges.
Reactivity over proactivity: Instead of thinking big and setting direction, you only react to what Claude outputs next. The tool dictates the agenda — not you. This is a creeping inversion of the leadership paradigm that barely anyone notices until the damage is too great.
Research from the field of Human-Computer Interaction, including work from Harvard and MIT, consistently shows: intensive technological dependency without recovery phases leads to measurable reduction in executive control functions — planning, prioritization, impulse control. These are precisely the cognitive capacities that leaders are paid for.
Leadership 2026: Leading AI with Rhythm
The answer to all of this is not to use less AI. That would be like telling an elite runner to stop training because they're overtired. The answer is periodization. Rhythm. Structured alternation of intensity and recovery.
Leadership 2026 therefore means: managing your own AI usage in a healthy way — and your team's. That's not softness. That's strategic self-leadership that secures high performance in the long term.
Concrete Rhythms for AI Usage
Establishing Rhythm in the Team
As a leader you bear responsibility not only for your own AI health but also for your team's. If you send prompts at midnight and that's expected, it shapes team culture. If you prioritize "AI use" over "results" and "recovery," you create a system that becomes collectively dysregulated.
Concretely: lead a team conversation about AI rhythm. Not as policy, but as genuine reflection — have you found a rhythm yet? What works? Where have the boundaries blurred? Leaders who lead this conversation build teams that perform with AI long-term — not teams that are burned out in six months.
My Own Approach
After the 3am moment I consciously rebuilt my AI rhythm. Mornings: no AI in the first hour. Instead notebook, coffee, slow thinking without output expectations. Then two intensive AI work blocks. Evenings: phone away after 9pm, Claude closed. Once a week I write my most important strategic thoughts without AI support — just to make sure I still can.
The result: I use AI more intensively and more effectively than ever before — because I no longer prompt out of habit, but out of decision. The rhythm didn't restrict the AI. It made it better.