The Computer & Technology Network

Stuart Armstrong’s *Smarter Than Us* warned about AI risk back in 2014. I read it in 2025 and was struck by how much of what he predicted is now happening — from AI agents to the moral philosophy problem. Here are my key takeaways, merged with my own experiences of using AI in 2026.

Concerned about AI weakening your cognitive abilities? Learn how to integrate AI with your PKM safely. This three-level framework establishes boundaries, protects essential thinking processes, and uses AI to enhance rather than replace cognition.

Levels 1 and 2 of my AI PKM framework teach you to use AI for critical refinement and as a thinking partner, whilst maintaining the cognitive friction needed to build genuine personal knowledge.

Here’s the challenge: you need cognitive skills to use AI effectively with PKM, but AI can undermine those very skills. This readiness assessment helps you determine if your PKM is ready for AI—or if you need more time at Level 0.

The AI cognitive ability paradox means beginners need to build skills before AI can help. Level 0 shows new Zettelkasten users what to automate and what to protect, ensuring AI strengthens rather than undermines your learning.

Much of our reality exists because we collectively agree it does. Technology has always shaped these shared beliefs. Now generative AI is creating content, not just transmitting it. What will this mean for our collective future?