This article explores emerging research and ideas about attention, technology, and digital environments. It is not intended as medical diagnosis or treatment advice.
Late at night. You've lost count of how many short videos you've watched, but your thumb is still moving.
One more refresh. One more recommendation. One more piece of content selected by a system designed to keep your attention.
The uncomfortable question is not only:
"Why can't I stop?"
It is:
"What kind of environment is shaping the way I think?"
In 2025, researchers published a study exploring what happens when large language models are continuously exposed to low-quality, high-engagement online content. The results were surprising: models trained on this type of data showed declines in reasoning, long-context understanding, and other capabilities.
The researchers called this phenomenon "LLM brain rot."
The comparison with humans is not literal. Human brains are not neural networks, and our minds are not trained in the same way AI models are trained.
But the analogy raises an important question:
If the information environment can influence artificial intelligence, what does our own information environment do to us?
The Silicon Mirror
The AI study is interesting because it reveals something about intelligence itself.
Large language models learn patterns from the data they receive. When that data becomes dominated by shallow, highly engaging, or low-quality patterns, the model's internal representations can shift.
The lesson is not that AI and humans work the same way.
The lesson is that inputs matter.
Every intelligent system adapts to its environment.
For AI, the environment is training data.
For humans, the environment includes:
- conversations
- books
- media
- interfaces
- social systems
- daily habits
The question is not whether technology changes us.
It is how.
The Disappearing Baseline
One of the hardest problems with attention is that we rarely notice gradual change.
A person who has always lived inside an environment of constant notifications and endless feeds may not feel that anything has been lost.
The baseline has moved.
This is why attention problems are difficult to measure. We often compare ourselves against yesterday, not against a different possible version of ourselves.
The concern is not that digital technology permanently destroys our ability to think.
The more realistic concern is that frequently practiced patterns become easier:
- rapid switching instead of sustained focus
- immediate feedback instead of delayed rewards
- consumption instead of reflection
The brain adapts to what we repeatedly ask it to do.
The Hidden Cost of Cognitive Offloading
AI introduces a new version of an old question:
What happens when we outsource too much thinking?
External tools have always expanded human ability. Writing, calculators, search engines, and computers all reduce cognitive load.
The problem is not using tools.
The problem is losing the ability to perform the underlying skill.
Recent discussions around AI in education highlight this distinction: AI can improve task performance, but completing a task does not automatically mean learning has occurred.
A student who receives a better answer is not necessarily a student who developed better reasoning.
The same principle applies beyond education.
Efficiency and capability are not always the same thing.
Building Cognitive Resilience
If digital environments influence attention, the answer is not abandoning technology.
The goal is building environments that protect important cognitive abilities.
Personal: Create attention preserves
Small boundaries matter:
- reading without notifications
- keeping phones away during deep work
- protecting moments of boredom
- spending time in environments that do not constantly demand reaction
The goal is not digital purity.
It is cognitive variety.
Educational: Protect the struggle of learning
AI should support thinking, not replace the process of thinking.
The most valuable skills are often developed through effort:
- forming questions
- making mistakes
- connecting ideas
- explaining concepts independently
Institutional: Measure more than output
Organizations increasingly measure productivity.
But productivity alone does not reveal cognitive health.
A better question is:
Are people becoming more capable, or only becoming faster at producing outputs?
Moving Forward
I think about the AI model trained on noisy data.
I think about people scrolling through endless feeds.
The comparison is imperfect, but the question remains:
What kind of environment are we building for intelligence?
The future is not about returning to a world before technology.
That world no longer exists.
The challenge is learning how to move between two modes:
deep focus and digital speed.
reflection and reaction.
tools and judgment.
The goal is not to escape technology.
The goal is to remain human while using it.
Continue Exploring: Reclaiming Your Attention
Attention is shaped by the environments we build around ourselves. Explore practical experiments and deeper perspectives on rebuilding a healthier relationship with technology.
-
48-Hour Offline Experiment: How to Break the Doomscrolling Loop and Reclaim Your Attention
-
Debug Your Attention: Why You Can't Stop Scrolling and How Grayscale Mode Helps
References
- Xing, S., Hong, J., Wang, Y., et al. (2025). LLMs Can Get "Brain Rot"! arXiv:2510.13928. https://arxiv.org/abs/2510.13928
- OECD. (2026). Digital Education Outlook 2026: Exploring Effective Uses of Generative Artificial Intelligence in Education. https://doi.org/10.1787/062a7394-en
- Kaplan, S. (1995). The restorative benefits of nature: Toward an integrative framework. Journal of Environmental Psychology.