← Back to Writing

Two Books Everyone Should Read in the AI Age: The Pathless Path & The Good Work

April 11, 2026 · 5 min read

Two Books Everyone Should Read in the AI Age: The Pathless Path & The Good Work

Today I want to recommend two books that I think are essential reading in the AI age: Paul Millerd's The Pathless Path and The Good Work.

From Rigid Jobs to Doing What You Love

Looking back, our work used to be incredibly rigid. In most cases, it was all about promotions, expanding influence, growing your team's impact, hiring more people, and getting a bigger compensation package. Going forward, I believe we should lean toward doing good work — doing things we're genuinely interested in — because that's how you'll write better AI/agent prompts. The reality now is that many talented programmers can single-handedly do the work of an entire team. So what you choose to work on becomes the most critical question. For ordinary people like us, the most important thing is to do what we're good at and what we enjoy. Of course, making money along the way is ideal.

The Pathless Path: Finding Your Own Journey

In The Pathless Path, Paul Millerd tells his own story beautifully: he once loved his job and worked incredibly hard. Then he got sick, and realized none of it had any real value. He spent nearly a decade finding his own path. He earns less than before, but he's much happier. I've read both of his books recently and benefited greatly.

I've now read The Pathless Path three times.

Making Peace with Yourself: Mindfulness and Wu Wei

A few months ago, I was deeply burned out and anxious. Recently, I've learned to make peace with myself. After practicing mindfulness meditation, things have improved significantly. The book also discusses the Daoist concept of wu wei, which I'd like to share:

"There's a phrase in Chinese, 'wu wei,' that describes how I felt. In English, its translation is 'non-doing,' but not in the sense of doing nothing. Non-doing is not about escaping anything or being lazy but instead refers to a deep level of connectedness with the world. The Chinese philosopher Lao Tzu wrote about this more than 2,500 years ago in the Tao Te Ching: 'Less and less do you need to force things until finally you arrive at non‑action. When nothing is done, nothing is left undone. True mastery can be gained by letting things go their own way. It can't be gained by interfering.'"

Wu wei is not about escaping or being lazy — it's about achieving a deep connection with the world. As Lao Tzu wrote in the Tao Te Ching over 2,500 years ago: "True mastery can be gained by letting things go their own way. It can't be gained by interfering."

— Paul Millerd, The Pathless Path

Flow State and Compromise at Microsoft

Trying to get colleagues to use AI at work is incredibly frustrating. You can never convince them. The existing compensation structure simply doesn't incentivize it. I think the future of programming will look more like sales — earning by commission. But defining that commission is extremely difficult. So at work, I've often found myself compromising too much. Looking back, my happiest time was my first year at Microsoft.

During that first year, I was basically working in a flow state. The Outlook iOS and Android email editor and renderer code was a total mess. Originally, Microsoft had acquired two companies: one called Sunrise and another called Accompli. The original team later left to work on Instacart, and I started cleaning up the codebase — converting a single JavaScript file into a TypeScript project and migrating from Browserify to Webpack.

But after that, I kept compromising — for immigration, for status, to earn a bit more to cover living expenses. I was never really happy.

"Roosevelt once argued: 'When you adopt the standards and the values of someone else or a community, you surrender your own integrity. You become, to the extent of your surrender, less of a human being.' I slowly learned this for myself, bouncing from job to job, doing the same thing Durant was doing, trying to achieve other people's goals. For years, I believed that once I reached some imagined leadership position in the future, I could truly be myself. This is an obvious illusion, but many people tell themselves this... I don't think I realized how much I hated it until I became self-employed and immediately stopped spending time hacking tests. As a freelancer, I competed on the quality of my ideas and my ability to deliver great work for clients. Many former consultants who become freelancers are amazed at how much less time it takes to do the same work. It's not because it's easier. In fact, without the resources of an entire company, it's harder. You just no longer need to impress hundreds of people."

— Paul Millerd, The Pathless Path

Deep Work vs. Shallow Work

Recently, I've realized it comes down to energy. When you're constantly interrupted by Slack messages and email notifications at work, you can't focus on deeper work — you can only do shallow tasks. This is exactly what Cal Newport describes in Deep Work: the shallower the work, the easier it is to replace. Now that we have AI agents, I believe the vast majority of shallow work can be fully automated. We should focus on our inner selves and do deeper work.

In traditional work environments — which is most people's reality today — you're constantly interrupted. That's what "collaboration" has become. If you can't escape constant interruption, you can't focus on prompting effectively, let alone orchestrate multiple agents simultaneously. I don't think the future of work will look like this, because AI can handle these tasks perfectly — and AI can be interrupted without any cost. Constant prompt steering actually improves AI's output. The more you interrupt AI, the better its output. But humans don't work that way. Humans and AI are fundamentally different.

Token Maxing Is Pointless

Recently, I've seen the "token maxing" trend — like Meta's dashboards where people compete on token usage. I think it's meaningless. The question isn't how much time you spend or how many tokens you use. The question is: as a person with agency, what do you choose to do?

The Good Work: Busyness as a Form of "Laziness"

"My friend and writer Noah Huisman has a unique perspective on this disconnect. He points out that this busyness can actually be seen as a form of 'laziness,' because 'laziness is not the opposite of diligence; it is the opposite of zeal... It is an indifference to truth, goodness, and beauty.' After understanding this inversion, I finally saw how I had spent a decade in an industry, doing massive amounts of work, yet feeling as if I hadn't created anything of real substance. For most of that time, I ignored the fact that I was being paid to do work while deliberately ignoring my own indifference toward it."

— Paul Millerd, The Good Work

So lately, I've been trying to limit my AI usage to four hours a day maximum, and spend more time exercising, reading, doing housework, and listening to podcasts. Even this writing — I'm doing it partly to offset the anxiety and false busyness of the AI age.

"When I slowed down, I could tell the difference between the work I truly cared about and the 'empty calorie' work — growth hacking, redesigning my website, tinkering with new productivity tools, and endless networking — all of which can be incredibly distracting on the entrepreneurial path."

— Paul Millerd, The Good Work

Final Thoughts

That's all for today. There's actually a lot more I want to write about, but I'll save it for next time. I believe that now that everyone has access to the leverage of Large Language Models, what you choose to do matters far more than it ever did before. The definition of "hard work" has been completely rewritten — and even the employment relationship itself is about to be rewritten. I hope this post helps.

Connect with me: