Deep Focus is a Skill and Skills Can be Trained

I wasn't born with the ability to do deep work; I developed it like any other skill.

Share
Deep Focus is a Skill and Skills Can be Trained
Photo by Ambitious Studio* | Rick Barrett / Unsplash

A mentee told me during our one-on-one session the other week that they were feeling a bit frustrated and demoralized. The reason isn't rare or unique these days, but the thought process basically goes like this:

  • I'm really excited to learn how to code and work with tech like a professional.
  • To meet my goals, I understand I need to work hard at learning.
  • I also understand that the friction of not understanding something is critical to mastering it.
  • But I can't seem to focus for more than a few minutes.

Now, I'm not trying to channel "old man blames new tech" vibes, because it's not like everyone could engage in deep focus before the internet either. It just feels more common these days. Even my wife, who is a bit younger than I am, makes comments about how I can just lose myself in work tasks for hours, barely moving, oblivious to the world around me. She doesn't get how I can do it.

Well, I have good news: I wasn't born able to do this. Nobody is. In fact, my Mother would tell you that I was a "hyperactive child" who could barely sit still.

Why Does it Matter?

Many IT roles, and software development in particular, involve a lot of abstraction. Languages, frameworks, systems, networks, protocols. Sure, the code can look simple, but when you are implementing a complex system, you must build a mental model if you want to be productive. "How does this function fit within the rest of the system?" "What are the things that could go wrong, and how should I mitigate the risk?" "If something is broken, how do I know where to look for the problem?"

Building that mental model so that you can keep the macro in mind while working on the micro requires deep focus. That is why, for years, developers fought for offices and quiet environments. Any serious distraction, such as a phone call, alert, instant message, or even just someone walking by your desk, can bring that mental model crashing down. Then you need to spend time rebuilding it.

This is one reason why agentic coding is exhausting to many developers. Having multiple agents run and complete work asynchronously creates an interrupt-driven workflow, which makes it nearly impossible to achieve a deep-focus flow state.

But that's a tale for another day.

Focus is Trained, Not Inborn

First, we must accept that sustained attention is not a personality trait. It's a skill built the same way as every other professional skill: by practice.

IT work does not yield to a TikTok attention span. You cannot debug a problem, hold a complex system in your head, or read through pages of log files and documentation by skimming them. So, when I hear someone say something like "I struggle with focus", that just means they haven't practiced focus enough.

The Good News and the Bad News

The good news is that you haven't lost anything, but this is where I will blame social media: you've spent years getting very good at something; it's just not the thing you need.

Short-form platforms optimize for engagement. Engagement means a tight loop of novelty and reward. Every swipe is a fresh dopamine hit; many videos have payoffs in under 3 minutes. This is how the product is deliberately designed, and doing it for hours a day, for years, trains your brain to become genuinely excellent at one specific skill: context switching.

This is one of the first things I learned about social media via my YouTube channel. You get about 10 seconds (the hook) to grab someone's attention and hope they stay for the rest of the video. By nature, I am not a hype person; I genuinely hate writing hooks.

Deep work is the opposite skill. Persevering in the stretch where nothing is rewarding yet, where the payoff is unknown, the material is hard, possibly frustrating, and your brain is begging to do literally anything else is genuinely difficult and uncomfortable.

That discomfort is a habit; it's not that there's something wrong with you. It's your brain trying to move you to that escape hatch where the dopamine comes easy. The moment you stop being entertained... you swipe and try again. It's so quick and easy that you stop thinking about it, and an hour later you've swiped dozens of times and only seen a few things worth looking at. But you've lost time, attention, and your ability to focus.

AI Makes Deep Focus More Important

"But if AI is producing more work, then focus doesn't matter as much, right?"

Wrong.

The way LLMs work today makes focus more important. These tools are genuinely useful, but they also quickly hand you confident, plausible-looking answers. Knowing whether that answer is the right approach is the skill that separates the professional from the "sloperator", as I like to call them.

On the B2B side of Skill Foundry's business, I train leaders and individual contributors in AI literacy and provide extensive advisory on where and how AI use is appropriate. So much of my AI work is grounded in the real world, where companies want to see a return on investment. This means I need to know what I'm talking about. It means I have spent over a thousand hours exploring the models, tools, and systems. And here are a few things I've learned:

LLMs are getting pretty good in a narrow context
What is the Linux command to do X? Provided it's something well documented and reinforced in the training data, the tools are almost always correct.

LLMs are hot garbage at architecture, strategy, and most things that involve critical thinking
It doesn't have real context; it doesn't understand what's already in place, and it happily fills in the gaps you leave in your instructions. In short, you need a capable operator to steer the tool and get value out of it.

And, if you want to achieve the "holy grail" of LLMs, the "one-shot prompt", you must have the skills to describe, in detail, exactly what you want. And how do you do that? By developing the skills to build the instructions.

As an aside, AI tools rarely one-shot anything without a lot of initial setup and structure. If I have to see another crappy CounterStrike knockoff demo, I'm going to lose it.

Evaluating AI Output is a Deep Focus Activity
The shine is starting to come off the speed arguments. Creating slop faster isn't economically valuable. There are more apps on the app stores than ever before, yet revenue hasn't gone up. Congrats, you made something fast... that no one wants.

And going too fast reduces quality. When you show me poorly performing software, it's always at least one leg of the trinity being neglected. What is the trinity?

  1. Fast
  2. Cheap
  3. Good

Pick two. You want fast and good? That isn't cheap. You want cheap and fast? It's not going to be good. Good and cheap? Possible, but it's slow.

I'm seeing this in the trenches right now. Bugs are slipping into production, and software quality is tanking. Beyond the fast-cheap-good problem, a big part of the decline in software quality is that reviewing code you didn't write is substantially harder than reviewing code you did.

Anyone who professionally does code reviews can tell you this. When you review code you didn't write, you have to understand the intent behind it, build the mental model, and then check the code. There are a lot of things I can write myself faster than I can read. Especially in my Skill Foundry, because I code-review learners on different projects at various points in the learning experience nearly every day.

And for the vibe coders: how do you build a mental model if you don't understand coding fundamentals and underlying systems? Simply put, you can't.

Training Deep Focus

Ok, let's get down to details. You know you need to train your brain, but how do you do it? Here are some tips:

Slow is Smooth, Smooth is Fast

I know my mentees are excited to learn to code. And when they sign up, their motivation is likely at its highest. So when we're setting initial goals, many of them are really ambitious. "I'm going to work 40 hours a week, every week, and learn really quickly!" Then, inevitably, they come back a week or two later and admit that they couldn't maintain their focus for that many hours.

This is similar to exercise. If you want to get really good at push-ups, you don't start by trying to do 100 unbroken. You start with whatever you can handle; maybe that's only 5. So you do 5 every other day until it's easy, then 10, then 12, then 15, and you incrementally increase the effort until you can do 100.

So, when you're getting started, figure out your tolerance for focus. Maybe it's only 20 minutes. That's ok; do 20 minutes per session, consistently, then extend.

Don't Fight Distractions, Remove Them

You don't get bonus points for sitting there, sweating, resisting the urge to check your phone all day. One reason I'm so successful at deep focus, besides practice, is that I literally remove distractions from my workspace.

  • I put my phone and computer notifications on "do not disturb". The phone stays in my pocket. Some studies show that even having your phone on the table can make you want to look at it. If you feel that strongly, go a step further and put the phone in a different room.
  • I have a separate computer for work vs. gaming and entertainment. So when I sit at the work computer, there's no temptation to play. (You don't need a separate computer; you can create a separate login account on yours that only has the applications installed you need for learning)
  • I only open the applications required for the task at hand. If I'm not answering emails, I turn off the email client. I'm not an EMT. No one is going to die if I don't answer right away.

Get Comfortable with Boredom

Maybe boredom is the wrong word. Get comfortable with stillness. The industry has trained you to reflexively fill every waking moment. Practice being still, take up meditation techniques, put on some white noise in the background, and just... be. Let your subconscious mind chew on the problem a bit.

Embrace Single Tasking

I don't care how good you think you are at multitasking. When you're configuring networks, writing code, or performing other skills that IT professionals get paid for, one missed command, one character out of place, and you end up with errors. Depending on what you're doing, the stakes can be really high, so you must be able to put away distractions and focus deeply on what you're doing and where you're going.

Protect the Flow

It takes time to get into the flow state where distractions and intrusive thoughts fade into the background. For me personally, it takes 15 to 20 minutes. Every interruption resets that clock. Defend uninterrupted blocks of time.

My wife and I share an office space, but I'm fortunate enough to also have a sound room that I use for my social media. As much as I love my wife, I'm better off working in isolation in the sound room if I need to do deep work, because she loves me and wants to talk to me, even when it's not a good time!

This can be one of the most challenging things for learners, particularly those with families. When and where can you find quiet time to focus? Make a space.

Be Patient With Yourself, and Take Breaks

You spent years training the switching muscle. Retraining takes consistent reps over time, and it'll feel bad before it feels good. Just as soreness signals your muscles are growing, boredom and dopamine withdrawal signal your brain is rewiring for deep focus.

But be kind to yourself and be patient. It isn't uncommon to take months of practice before you start feeling comfortable sitting down and working into a flow state.

The Takeaway

Back to my mentee, and to you if you saw yourself in any of this. The struggle isn't evidence that you can't do it. It's evidence that you haven't trained it up yet. Discipline, patience, and a willingness to embrace discomfort are what it takes to succeed.