My daughter and I were hand in hand as we set off on the mile-long walk to school. She is nine. Her hand is still small enough to disappear in mine, though not for much longer. One-on-one time with her is rare now. Her little brother is almost seven, and he fills whatever room he is in. We took the mile slower than usual.

She talked about gymnastics most of the way. She was eager to tell me which skills she could do with ease, and which she was close to, and how close, down to the inch. Somewhere in the middle of the third accounting of her kickovers, she stopped talking. I looked down at her. She was thinking. After a moment, she looked up and asked, “What would be a good job when I grow up?”

I smiled in relief. I had been bracing for a conversation about how babies are made. It took me a few seconds to realize I had not gotten the easier question.

Five years ago, I could have offered half a dozen respectable answers without much trouble. My parents’ line was that the world always needs doctors, lawyers, engineers, and trash collectors, and since nobody wants to be a trash collector, I had better do well in school. That advice has not stopped being true, but the world will need far fewer of all of them.

The supply chain analyst job I took right out of school is being done by AI now. So is a growing share of the work that doctors, lawyers, and engineers do. Even the trash truck is starting to drive itself.

Birthrates across the developed world are collapsing, and for the first time in modern history, most of humanity lives in a country where each generation is smaller than the last. The cheap energy, abundant materials, and open borders I grew up with are all being renegotiated at once: wars and datacenters resetting the price of power, minerals turning into matters of national security, tariffs returning, borders closing.

Each of these would be a generational shift on its own. Yet none of this adds up to the end of the world. It just means the work we do today will not be what we do in twenty years.

As I walked beside her, the best I could offer was this: stay flexible. Learn how to learn, because what you do first will not be what you do at forty. And learn how people work. Machines change, but people are people, and have been for several hundred thousand years. Whatever you end up doing, you’ll do it with other people, and that understanding will always be in demand.

I heard how thin it sounded.

---

What I told her was less fatherly wisdom and more a description of my own life. I came of age on dial-up internet. By the time I was starting companies, the frontier had moved to messaging, then to smartphones, then to the cloud, and now to AI. My four decades have spanned one of the fastest stretches of change in human history. So many “once-in-a-lifetime” events have come and gone that I stopped counting.

I never knew what job I was going to do. How could I? The terrain kept changing. There was only opportunity. I followed where it pulled and taught myself whatever came next.

The pattern started at sixteen. I was rebuilding apartment buildings gutted by meth explosions and patching up crack houses. I routinely found heroin needles in the cupboards, not far from the kids’ cereal.

At seventeen, wanting to work in better neighborhoods, I started a high-end home remodeling business with skills I had taught myself off the internet (pre-YouTube). My clients were the doctors, lawyers, and engineers my parents told me to emulate. It paid my way through university.

At twenty-three, I saw a news report about copper reaching all-time highs. I proceeded to mine 50 tons of pre-1982 U.S. copper pennies out of circulation, each one worth more as scrap metal than as money, and created nine different ways to sell a penny for more than a penny, earning enough to fund my MBA.

At twenty-six, with the MBA in hand and mobile phones spreading faster than banks could follow, I started my first tech company, banking the unbanked in Mexico and Colombia. We were the first to do cross-border remittances over SMS.

At twenty-nine, I was granted patents on technology that today underpins nearly every mobile banking transaction in the Western world. Hundreds of millions of people rely on it daily.

At thirty-six, during the Covid-19 lockdown, I sold the global communications company we built. The remote-work revolution ran on our platform. Twilio, MessageBird, RingCentral, Microsoft, and others were customers.

At thirty-seven, with the sale closed and something new gathering speed, I started an AI company, six months before the launch of ChatGPT. At the end of 2025, as we neared profitability, I shut it down and refunded every customer. What the market wanted and what I was willing to build had diverged too far.

Now at forty-two, something is pulling at me again.

---

We reached the school gate, the bell rang, and I hugged her goodbye. Walking slowly back toward home, I knew that my answer dodged the real question. Everything I had told her assumes she can tell what is real. That the teachers she learns from are real. That what she learns is real. That the people she comes to trust and rely on are real.

When I was her age, none of this was in question. Mrs. Bertalone taught me geography and politics. I learned biology from Mr. McDonald. The books they handed me were written by humans who put their names on them. Even the bullies in the schoolyard had a physical nose I could bloody when they ran their mouths. Bullshit artists existed, but they could not reach me at scale.

When more and more of what she sees is made for her, more and more of her life happens outside the physical world, and more and more of who she meets may not be anyone at all, will she be able to tell what is real?

The question would not leave me on the slow walk home. I had walked away from my last business without being able to name what had made me do it. Not fully. But now I could.

The market wanted only two things from AI: help me reduce the number of humans I employ; help me capture more of my audience’s attention. I had no answer for how the average person would be better off with either of these forced upon them. While businesses were asking me to engineer phony engagement at unfathomable scale, I had built the company around the opposite: using AI to make work more human, handing the machines the mundane and giving people back meaningful work. The companies winning were the ones who said yes — flooding the world with synthetic, unreal, trust-eroding bullshitBullshitverb.To choose to act as though you care about the truth when you do not..

It is today’s winners who set the conditions for the world my daughter will grow up navigating. Walking away did not slow any of it down. The economic incentives are too strong for anyone inside the system to stop it. This is what is next. My daughter will come of age in the era of bullshit whether I like it or not.

All I can do is prepare. Myself first, so I know what I am looking at. Then her.

Figuring out what is real is slow work. It requires first-hand contact with the world: reading old books, working with my hands, testing what I think against what you think, being wrong out loud, and correcting course. And honest, thoughtful dialog, not the parody of it that fills comment sections.

This is what a good classical education used to do. It is what a good conversation still does. And it is exactly what the current shape of the internet and media landscape is engineered against.

This is what True to Things is for. Not a feed. Not a weekly newsletter. A place to think and learn in the open, long-form, grounded in real world experience and the best that has been written on the subject, in dialog with other people trying to do the same.

Someday soon, my daughter will ask me again. I cannot dodge her a second time.

Matt
matt@truetothings.com