# Resistance in Working with AI: How Not to Lose Cognitive Skills
AI simplifies tasks by removing cognitive resistance, leading to risks of diminished critical thinking, creativity, and focus. The author shares observations and strategies for maintaining agency in the LLM era.
Risks of Over-Reliance on AI
Studies show a correlation between frequent AI use and declining cognitive abilities. Critical thinking weakens: people overlook inconsistencies in algorithm recommendations due to automation bias. Decision-making skills suffer, as delegating tasks to AI reduces practice in analysis.
Creativity drops—it's easier to ask GPT for 10 variants than to generate ideas yourself. Memory deteriorates from offloading memorization to machines, though the internet already weakened it before. Focus gets shattered: AI's short answers create an illusion of understanding, like TikTok for knowledge.
Historical analogies confirm the pattern. Calculators diminished mental math skills, GPS reduced hippocampus activation, the internet changed how we process facts. Offloading mental load to technology boosts efficiency but causes skill atrophy. Paul Graham warned: over-reliance on tools will destroy the ability to write.
However, AI can complement rather than replace humans, by redistributing cognitive load to new skills.
Key Limitation of AI: Lack of Resistance
ChatGPT nailed it: AI removes the resistance needed for growth. In complex tasks, struggling to formulate thoughts activates deep thinking. Searching for words in music hones taste, reading nonverbal cues in meetings builds empathy.
Writing an article demands observations, reading, discussions—the process sparks insights and satisfaction. Without resistance, agency slips away.
Dangerous AI usage scenarios:
- Replacing thinking: writing texts without truly understanding them.
- Ignoring context: AI misses nonverbal signals (less than 10% of information is verbal).
- Avoiding difficulties: summaries instead of deep dives.
- Loss of pleasure: automating hobbies kills the joy of creation.
Strategies for Preserving Cognitive Agency
Main rule: think independently before consulting AI. Formulating the problem kickstarts your thinking, even if the prompt ends up simplified. Prompt engineering is evolving—now it trains the human, not just the model.
Practical recommendations:
- Analyze independently: gather context and formulate hypotheses before querying.
- Use AI for verification: after drawing your own conclusions, refine with the model.
- Train resistance: write drafts by hand, hold meetings without AI prompts.
- Monitor skills: track where AI is replacing practice.
- Balance the load: delegate routine tasks, but keep creativity and analysis for yourself.
This shifts skills: from routine calculations to strategic thinking and interpretation.
What Matters
- Automation bias undermines critical analysis—always verify AI recommendations.
- Resistance is essential for deep thinking and empathy.
- Think for yourself first: it preserves agency.
- AI redistributes skills rather than dumbing you down—adapt mindfully.
- Historical technologies prove it: skills atrophy without practice.
— Editorial Team
No comments yet.