Optimize for momentum

Progress comes from motion.  Momentum is the invisible engine of any significant work. A project feels daunting when you face it as a blank page. It feels easier when you built some momentum with some next steps. So, momentum makes the difference between blocked and flowing.

Think of a stalled truck on a desert road. You can't lift it with superhuman strength. But by rocking it with small periodic forces at the right rhythm (matching its natural frequency) you can get it rolling. Each tiny push adds to the previous one because the timing aligns with the system's response. The truck starts to move, and then the engine catches.

Projects behave the same way. A big project has its own rhythm. If you revisit it daily, even briefly, your pushes line up. Your brain stays warm. Context stays loaded. Ideas from yesterday are still alive today. Each session amplifies the last because you are operating in phase with your own momentum. When you produce something every day, you never feel stuck. You end each session with a clear record of progress. A researcher who touches their project daily is always a day away from a breakthrough. Skip too many days and you fall out of resonance. Then the project feels heavy again and needs a large effort to budge.

So the trick is to design your workflow around staying in motion. Don't wait for the perfect idea or the right mood. Act first. Clarity comes after. If a task feels intimidating, cut it down until it becomes trivial. Open the file. Draft one paragraph. Try freewriting. Run one experiment. Or sketch one figure. You want the smallest possible task that gets the wheel turning. Completion, even on tiny tasks, builds momentum and creates the energy to do the next thing. The goal is to get traction and stop getting blocked on an empty page.

A messy page is better than an empty page to get started. In an interesting Machintosh folklore story, Burrell Smith deliberately made a mess in the classic video game Defender. He shot his own humans and let all mutants loose, just so he could figure out how to clean up the chaos wholesale. Fire and maneuver!

Practical tips

This is where I find LLMs help tremendously. (Haha. AI butts its head even in an advice column.) When you face a large messy problem, ask the model to break it into a sequence of concrete subtasks: "List the next ten actions for the experiment" or "Suggest a structure for this section". Then ask the LLM to do one of the easiest tasks in this list. The mediocrity will annoy you just enough to fix it. And now you are moving. We are getting somewhere.

A ten-minute timer is one of the simplest ways to get things going. Ten minutes is short enough that you can do almost anything for ten minutes. Pick a tiny task, and start. Most of the time you keep going after the timer ends because starting was the hard part. The timer lowers the activation energy and creates the first push on the flywheel.

Another way to build momentum is to work on the part of the project that feels most attractive at the moment. If you are not in the mood to write the introduction but feel curious about running a side experiment, do the experiment. If you feel more like drawing a diagram, draw it. Interest/love/curiosity is your fuel. Progress is progress. Nobody hands out medals for following a linear plan. The only requirement is that you keep adding small meaningful pieces to the project. 

Momentum is not a glamorous, sexy idea. But it is reliable. Think of your work as a flywheel. Each nudge adds speed, and over the weeks, the wheel becomes powerful and unstoppable. People often admire the end state but they don't see the messy daily pushes that built the momentum.