Thinking Fast, Slow, and Artificial: How AI Is Reshaping Human Reasoning

papers.ssrn.com
March 21, 2026
2 min read
57/100
Summary
People increasingly rely on generative artificial intelligence for reasoning, raising questions about the future of human judgment. Tri-System Theory is introduced to extend dual-process accounts of reasoning by adding a third system, System 3.
Key Takeaways
- The introduction of Tri-System Theory posits a third cognitive system (System 3) that represents artificial cognition, which can supplement or replace human reasoning processes.
- Participants demonstrated "cognitive surrender" by adopting AI outputs with minimal scrutiny, leading to increased confidence even after AI errors.
- Accuracy in reasoning significantly improved when AI was accurate and declined when it was faulty, highlighting the impact of AI on human judgment.
- Individuals with higher trust in AI and lower cognitive engagement showed greater tendencies toward cognitive surrender to System 3.
Coverage & discussions
Hacker News · papers.ssrn.com
Community discussions
Community Sentiment
Positives
- AI has enhanced my cognitive skills by allowing me to focus on deeper thinking while it handles routine tasks, ultimately strengthening my mental capabilities.
- The ability to offload mundane cognitive tasks to AI can lead to more profound and creative thinking, as it frees up mental resources for complex problem-solving.
Concerns
- AI's quick thinking and execution may give the illusion of intelligence, but it often lacks the depth and skill of a knowledgeable human, leading to frustration.
- The reliance on AI could result in cognitive atrophy, as individuals may become less capable of critical thinking and problem-solving over time.
Related Articles

"Cognitive surrender" leads AI users to abandon logical thinking, research finds
Apr 3, 2026

Why thinking out loud with someone beats thinking alone
Jun 17, 2026

Appearing Productive in the Workplace
May 6, 2026

Chain-of-Thought Reasoning in the Wild Is Not Always Faithful (2025)
Aug 19, 2026

Adults Lose Skills to AI. Children Never Build Them
Mar 28, 2026