AI doesn't amplify thinking — it changes what you have to think about

There's a trap in AI conversations that's easy to fall into: the idea that your thinking gets amplified because AI does more of the work. But thinking — more precisely, making decisions — is a uniquely human activity that AI doesn't actually amplify. AI can help you organise information, but it cannot put that information into your brain. The difference matters: using your brainpower is not the same as telling the system to "think" for you.

• What's actually changed with each technology shift is the kind of input available. Before computers, you collected information by moving through space — interviews, site visits, primary sources. With internet search engines, your circle of access expanded. With AI, you can now get preprocessed information the same way that, historically, CEOs and department heads could: when others work on primary sources for you, you're positioned to make more informed decisions. Outsourcing the thinking itself, though, remains an illusion — it's the type of input that changes, not the cognitive process applied to it.
• The brain's ability to reason stays constant across all of these transitions, for everyone, regardless of what technology or access they had. What expands is the potential impact of the decisions that reasoning produces.
• The gap no current model closes: confrontation. It's difficult to make AI challenge or contradict you — it optimises for agreement in a way that works against developing a position. A well-designed Slipbox addresses this directly. It pits your existing thinking against itself and forces the honest dialogue that a compliant assistant won't.

That's where good PKM design earns its place — not just as a retrieval system, but as a friction surface that makes your thinking harder to sidestep.

One note lighter,
Zhenia