Getting Into Flow

I’ve been an avid listener of The Knowledge Project for the last several years, and I really appreciate the work Shane Parrish puts into it. So I’m writing about a recent episode: Steven Kotler: The Science of Doing Your Best Work. Kotler is the executive director of the Flow Research Collective and has spent decades studying flow states.

I produce my best work when I’m in flow. Before this conversation, I thought that flow state was a mindset that you subconsciously transitioned into while performing some task. Kotler makes the case that there are deliberate actions that we can take to make getting into flow state more likely, and so I thought I’d write a short post on them.

Quick disclaimer: I’m not a scientist. I can only speak as a programmer, and for me, getting into flow on demand is one of the most sought-after skills there is. Unless I link something else, everything below is Kotler’s claim. Where I couldn’t find a source, I say so.

Flow follows flow

Kotler talks about a primary flow activity: something you’ve done since you were younger that you tend to get completely absorbed in. For him, it’s skiing. For others it’s dancing or playing an instrument. The most common activity is reading. For me, I think it’s exercise or walking in nature.

Flow is a focusing skill, so time spent in flow on Sunday makes it easier to get there on Tuesday. His idea is that spending time in flow outside of work can make it easier to find flow at work. He suggests two to four hours a week in that primary activity.¹

He also says that if you’re burned out, doubling down on your primary flow activity is the way back, which I found to be particularly surprising. When you’re burned out, it seems like getting into a flow state would be on the opposite end of the spectrum. I did a little bit of digging and the research I found is a bit more modest. A 2022 systematic review of 18 studies found flow and burnout symptoms tend to be negatively associated, so flow looks protective, though the authors call the findings mixed and want more work on why.

Stretch without snapping

A mentor once told me that you learn the most just outside your comfort zone and that has always stuck with me, so I was excited when I heard Kotler tie this into the flow conversation.

Flow shows up when a task is just a little harder than your current skill. This is the old challenge-skill balance idea, which has been revised over the years. Kotler puts a number on it: 4%.

So you want to stretch, but not snap.

He’s upfront that 4% is a back-of-the-envelope figure and not science. He’s also called it arbitrary and said it changes for every person. It lines up with the Yerkes-Dodson law, where performance rises with stress up to a point and then falls off. Too easy and you’re bored. Too hard and you’re anxious.

Design your environment

First, get rid of distractions. Kotler’s version: phone off, one channel open for emergencies, batch your messages.

I recently bought a timed lockbox for my phone and it’s been a game changer. I used to toss it somewhere random in the house and forget about it. Now it goes in the box, gets locked for a few hours, and I’m good to go.

Kotler also says to clean your desk. His reasoning is that a tidy space lowers cognitive load, and the brain redirects that energy to attention.

Building your flow capacity

Kotler says sleep runs in cycles of roughly 90 to 110 minutes, and your awake and alert cycle is about the same length. The idea traces back to Nathaniel Kleitman’s basic rest-activity cycle, and a 1994 study found cycles of 90 to 100 minutes in task performance and EEG. The waking version is still contested though, so I’m treating 90 minutes as a good starting point and not a law of nature.

You don’t need to start there.

Start by starting.

If you can only focus for 10 minutes, start with 10. Then 15, then 20. Kotler says people who stick with it are surprised when they suddenly hit 90. I’ve started planning my days around 90-minute blocks since coming into this information, and it’s working for me.

Find the part of the task that matters to you

Kotler argues that ADHD is really “hyperflow”. The problem isn’t focusing, it’s focusing on things you don’t care about - a scenario I know all too well.

The science here is less settled. A recent review says researchers still disagree on whether hyperfocus and flow are the same thing. I won’t say much more on it, but the reframing tip stuck with me.

Kotler says attention and autonomy are tied together, so you have to find the part of a task that’s yours. His example is an early journalism assignment about data caves. He didn’t care about data caves, but he did care about learning how Wired structures its stories.

The developer version: maybe I don’t care much about AWS itself, but I do want to understand how to set up a load balancer. That gives me a way into the task. It’s one idea to try, not an explanation for every person’s experience with ADHD.

What’s happening in your head

During flow, Kotler says most of the prefrontal cortex turns off as an “efficiency exchange.” The brain shuts down noncritical structures and puts the energy toward focus and attention. The name for this is transient hypofrontality, proposed in Arne Dietrich’s 2004 paper.

The imaging work is more nuanced than “it turns off.” A 2014 fMRI study of 27 people found less activity in the medial prefrontal cortex and amygdala during flow, and more in the left inferior frontal gyrus and putamen.

Kotler also says researchers look at chemicals like norepinephrine, dopamine, serotonin, anandamide, and endorphins. His takeaway is that flow is a measurable state and not an imaginary concept, and measurable things can be trained.

Play beats repetition

Kotler compares deliberate practice (repetition with small improvements) to deliberate play (repetition with improvisation). Outside tightly bounded fields like math or violin, he says play wins, because it brings more neurochemicals and those tag an experience as worth saving.²

As a software engineer I can read the docs for a new library, or I can build a project with it. I’m betting on the project. Building will teach me way more than reading will.

The short version

Clear your distractions and your desk.
Pick work that sits a little outside your comfort zone.
Start with whatever focus block you’ve got and build toward 90 to 110 minutes.
Make time for your primary flow activity. Mine is going for a walk.

This is the number Kotler gives in the episode, and I couldn’t find the study behind it. A 2021 Psychology Today piece puts his weekly guidance at two to six hours, so treat the exact range as a rough guide. ↩
Kotler says his own lab compared the two, but I couldn’t find a published paper. The closest I found is a Flow Research Collective Radio episode where he talks about the distinction. Take it as his claim for now. ↩