Why I Stopped Trying to Pick the Right Career for My Daughters
The question came from the back seat, somewhere between home and her public speaking class.
“Daddy, what time is late morning?”
She’d asked because we had a playdate the next day, and I’d told her it was late morning. She wanted to know when to expect it.
My instinct was to say 10, maybe 11. Done in two seconds.
Instead I asked her what she thought morning was. She said 7 to 9am. I asked what came after 9. She went quiet for a moment, then offered “late morning?” like she wasn’t sure she was allowed to answer her own question.
She was right. She just didn’t know it yet.
Our daughters are seven. Old enough to argue, not yet old enough to care about the future of work.
So when I think about AI and what it might mean for them, I am not having that conversation with them. I am having it with my wife.
It happens in fragments. After a school event. While reviewing homework. On the drive back from enrichment classes.
Because I work in AI, I feel a quiet pressure to optimise early. Should they learn to code? Should they be using AI tools already? Are we late?
I am increasingly skeptical of that frame.
Every few months there is a new wave of predictions about which jobs AI will take and which will survive. I find myself less interested in that conversation.
The timelines keep shifting. The categories keep changing. By the time our daughters enter the workforce, the jobs we are debating today may not be the relevant ones.
So my wife and I have been trying to focus on something more durable. Not which job. But what kind of thinking.
A few weeks ago, I was telling my wife I was on a strict diet. One of our daughters overheard, and at dinner that night she announced she was on a strict diet too and pushed her food aside, barely touched.
I could have just corrected her. Instead I asked what she thought “strict” meant. She thought about it. Then I asked whether not finishing her normal portion was the same as what Daddy was doing. She worked it out herself.
She didn’t need the answer. She needed the right questions.
This is what we keep coming back to. Getting comfortable finding answers rather than waiting for them. Sitting with not knowing, long enough to think.
When they ask something we know, our instinct is to answer immediately. We are trying to resist that. Sometimes we say “what do you think?” Sometimes we say “how would you find out?”
It is slower. Sometimes annoying for everyone involved.
But the world they are growing into will produce plausible answers on demand. Knowing how to navigate uncertainty, to stay steady when you are not sure, feels like it will matter more than any specific skill we could sign them up for.
I am not certain this is right. There is a part of me that wonders if we are underweighting technical fluency: whether early exposure to tools will compound in ways we cannot predict.
But I suspect many parents will rush toward the visible technical skills. And the scarcer advantage will remain judgment, curiosity, and the ability to ask a better question.
That is less dramatic than most AI forecasts.
It also feels more durable.