Lena Park

Model collapse is real, measurable, and already happening in the wild. Here's what it looks like mechanically and why it matters.

Starting fresh feels productive. It almost never is. The real leverage is in closing out what's already in motion.

A look at how Basecamp restructured around unscheduled communication and what it reveals about the hidden cost of synchronous coordination.

LLMs generate code by predicting plausible text, not by reasoning about execution. Understanding that gap changes how you use them.

Completion feels like progress. But a finished list just means you did everything you planned, not that you planned the right things.

Every line of code you keep is a liability. The best engineers know that removing code is often the highest-value work they can do.

AI models don't hedge at the edges of their knowledge. They extrapolate smoothly and state the result as fact. Here's what's actually happening.

Most task managers are optimized for capture, not completion. The design choices that make adding tasks effortless are the same ones making finishing them harder.

Getting better at productivity systems often produces more tasks, not fewer. Here's the mechanism, and what to do about it.

Canceling a bad meeting feels productive. But the real problem is that you scheduled it in the first place, and probably will again.

AI writing tools are getting genuinely impressive. That's exactly why they're quietly degrading the cognitive skill they're supposed to support.

Softmax converts raw model scores into probabilities. But what it actually does to those scores in the process is stranger and more consequential than most explanations let on.

Async-first isn't about time zones. It's a different theory of how thinking work actually gets done, and most offices still haven't figured it out.

Code-generating AI models predict text that looks like working code. That's not the same as knowing whether the code works.

The design of most task management systems quietly rewards capturing work over finishing it. Here's what that costs you and how to fix the incentive.

A product team's accidental experiment revealed that clearing calendars did more for output than any sprint ritual. Here's what actually happened.

Between your words and the model's attention sits a layer most users never see. Understanding it changes how you work with AI.

Canceling a meeting isn't laziness or avoidance. It's often the most productive decision you can make for everyone in the room.