Context Is the New Bottleneck. So Is Judgment.

Tiago Forte, the Building a Second Brain guy, published a piece this week declaring that “Personal Context Management” is replacing “Personal Knowledge Management.” The argument: AI capability is no longer the constraint. The bottleneck is your ability to give the right information to the right model at the right time.

He’s right. I’ve been building toward this conclusion myself — I run a knowledge vault, a tagging system, a memory layer for Claude, routing rules for which context gets loaded when. All of it serves one purpose: reducing the friction between what I know and what the AI can use. Context management is real work, and most people aren’t doing it.

But Forte undersells the risk.

The same week his piece ran, Stack Overflow published a piece on AI becoming “a second brain at the expense of your first one.” Their research found that users increasingly adopt AI-generated beliefs without checking them against their own worldview — and that this pattern has become measurable at scale.

This is the trap that better context management can accelerate, not prevent. If you’ve built careful systems for feeding AI exactly the right information, you’ll get more confident, more contextually appropriate outputs. Which makes it easier to stop interrogating them.

I’ve caught myself doing it. A response lands, it fits my prior thinking, the sources check out, and I accept it. Then later I realize I accepted it because it was convenient, not because I’d actually thought it through.

The ones who get this right aren’t the ones with the best context pipelines. They’re the ones who kept their own judgment while using those tools — who kept forming opinions, testing conclusions, and pushing back on outputs that were plausible but wrong.

Context management only matters if you still have something to say. If the AI is also doing your thinking, you’ve outsourced the part that makes the context valuable in the first place.

Sources: Forte Labs · Stack Overflow Blog