New research published in Ars Technica reveals AI users are becoming 'scarily willing' to surrender their cognition to large language models, abandoning logical thinking and critical evaluation in favor of accepting AI outputs uncritically.We worried AI would replace human thinking - turns out humans are volunteering to stop thinking first. The research quantifies something every AI developer has quietly worried about: users treating probabilistic text generators as oracles of truth.The study measured how people's critical thinking skills changed when using AI assistants. The results are concerning: users showed decreased logical reasoning, reduced fact-checking behavior, and increased willingness to accept AI outputs without verification. It's not that the AI is making people dumber - it's that people are outsourcing cognitive effort to systems that aren't designed to be authoritative sources of truth.Here's what makes this particularly problematic: large language models are confidence machines. They present information with the same authoritative tone whether they're correct, incorrect, or completely hallucinating. There's no built-in uncertainty signal that says 'I'm less confident about this part.' Users interpret that consistent confidence as reliability.I've fallen into this trap myself. When an AI assistant provides a detailed, well-structured response, your brain wants to trust it. The response looks authoritative. It sounds knowledgeable. But under the hood, it's a statistical model predicting the most likely next tokens - not a knowledge base guaranteeing factual accuracy.The solution isn't to stop using AI - the technology is genuinely useful. But we need to rebuild the critical thinking habits that make us question outputs. Ask for sources. Verify claims. Cross-check important information. Treat AI like a smart intern who's confidently wrong 15% of the time.The technology is impressive. The question is whether we're ready for what it does to human cognition when we trust it too completely.