Gen Z grads aren’t doing it wrong. The job market has changed.
What two words can help?Craft = the skills you build and own
Need = what people or the market actually value
A recent New York Times piece shows many grads are doing everything “right”—
degrees, applications, tailored resumes—
and still struggling to land jobs.
The process feels more competitive, less predictable, and harder to navigate than expected.
For many, the transition from school to work is taking longer—and proving more difficult—than they anticipated.
Young people are in a miasma of gloom about work. The AI threat, AI interviews (yup, they're being interviewed by machines), and the hiring numbers. You know. My question: what is this generation supposed to actually *do* in the face of this fear? When so much feels broken, how do you start?
Students, graduates: here’s how I believe you can give your dreams a fighting chance:
Don’t just look for a job. Establish a craft. An expertise or skill, built over time, that protects you from being treated as disposable or interchangeable. Craft guides the hand of the surgeon, restoring an accident victim’s body. It’s how a director holds an audience’s attention for hours. Why restaurant meals taste so good– and also why the best-written home recipes turn out beautifully in amateur hands. Any employee can be fired at any time, but your craft can never be taken away from you. And this is hard to see at age 22, but mastery of craft feels *amazing.*
Pair it with a need. *Enough* with the conventional wisdom that says you must study X to win. (When I was in high school, it was Japanese, because Japan was projected to take over the world economy. Instead, their stock market slumped for 30 years.) Every generation has a version of: Here’s the path for winners. Don’t get left behind. Hurry up and learn genetics. Study Mandarin. No, computer science.
Great pursuits, sure, but not golden tickets. The wiser way to seize the future is to think about need. What is your own assessment of what society will need most during your working years? In terms of care, products, and information. I’m fifty, and the people I see thriving at work chased a need—articulated through independent observation, not conventional wisdom.
Most important: Do not give up before you even start. Frustration and disappointment are certain. Failure is possible. But if you abdicate the search for satisfaction now, you will put it further out of reach. Resist the urge to arm yourself with uninformed cynicism masking as oh-so-wise pragmatism that’s really just good old fear of rejection. We do not yet know what the world will offer you.
I have covered the workplace for years (including breaking the Weinstein story with Megan Twohey), I have a 20-year-old daughter, and I can’t stand to see us give up a collective sense of aspiration. So I wrote a book called How to Start, from which this essay is adapted. Thanks for checking it out. Starting has never been easy. This era is making it harder. We need to bring young people all the help we can.
There’s definitely a sense of uncertainty right now, especially with how fast AI is changing the landscape.
I love 💓 the idea of focusing on a craft. To me, a craft is something you’re naturally good at, can do with ease, and can potentially get paid for.
In today’s results-driven world, being able to do something faster, better, and actually enjoy it, that’s powerful. 💪
Developing a craft feels like a more grounded and practical way to move forward, even when everything else feels unclear.
I love 💓 the idea of focusing on a craft. To me, a craft is something you’re naturally good at, can do with ease, and can potentially get paid for.
In today’s results-driven world, being able to do something faster, better, and actually enjoy it, that’s powerful. 💪
Developing a craft feels like a more grounded and practical way to move forward, even when everything else feels unclear.
The rise of an AI tool called Colleague Skill, which apparently lets coworkers "distill" colleagues' skills and personality traits and copy them into an AI agent, has ignited debate about the impacts of AI on workplace dynamics. The tool has gone viral across Chinese social media, generating humor and discomfort, according to MIT Technology Review. Ultimately, it raises concerns about employee dignity and the extent of automation's potential. While companies could benefit from understanding workflows better, experts say these "blueprints" could alienate workers by oversimplifying their contributions.
A product manager in Beijing built an open-source tool that sabotages AI agent training. Light mode, medium mode, heavy mode. You pick based on how closely your boss is watching.
The backstory: companies in China are telling employees to document their workflows so AI agents can replicate them. A GitHub project called "Colleague Skill" went viral for doing exactly that, claiming it could "distill" a coworker's skills and personality into an AI stand-in. Within days, a counter-tool appeared that rewrites the documentation into vague, non-actionable language. Useless to the AI. Plausible to the manager.
MIT Technology Review published the full story this week.
Any organization rolling out AI agents is going to bump into some version of this. If people believe the point is to replace them, they're going to protect themselves. They'll comply on the surface and hollow out the substance underneath. You end up with documentation that looks thorough but teaches the AI nothing.
That PwC data I posted about last week told the same story from a different angle. The companies pulling away weren't the ones spending the most on AI. They were the ones that were straight with people about what was changing and why.
Nobody can promise the job will stay the same. But if you need your team's knowledge to make AI work, tell them what's changing and where they fit.
The backstory: companies in China are telling employees to document their workflows so AI agents can replicate them. A GitHub project called "Colleague Skill" went viral for doing exactly that, claiming it could "distill" a coworker's skills and personality into an AI stand-in. Within days, a counter-tool appeared that rewrites the documentation into vague, non-actionable language. Useless to the AI. Plausible to the manager.
MIT Technology Review published the full story this week.
Any organization rolling out AI agents is going to bump into some version of this. If people believe the point is to replace them, they're going to protect themselves. They'll comply on the surface and hollow out the substance underneath. You end up with documentation that looks thorough but teaches the AI nothing.
That PwC data I posted about last week told the same story from a different angle. The companies pulling away weren't the ones spending the most on AI. They were the ones that were straight with people about what was changing and why.
Nobody can promise the job will stay the same. But if you need your team's knowledge to make AI work, tell them what's changing and where they fit.