I am watching my friends who spent $100k on specialized Master’s degrees slowly lose their minds. These are people who spent a decade becoming the ‘best’ at one specific thing—SEO backlink auditing, technical copywriting for medical devices, or writing specific types of Python scripts for legacy databases. They were the safe ones. They were the ones who followed the rules. And now, they are the ones getting the ‘we’re moving in a different direction’ emails on a Tuesday morning.
Specialization is becoming a trap. It used to be a fortress. You’d dig a moat around your niche, and nobody could touch you. But AI doesn’t care about your moat. AI is a specialist by nature. It can go deeper, faster, and cheaper than any human can in a narrow lane. If your entire value proposition is ‘I know this one specific thing better than anyone else,’ you aren’t an expert anymore. You’re a legacy system waiting to be decommissioned.
The only people I see who aren’t sweating right now are the weirdos. The ‘T-shaped’ people. They have that one deep anchor—maybe they’re great at sales or they actually understand how a supply chain works—but they have this broad, messy horizontal bar of skills that allows them to connect dots that the machines can’t see yet. They are the generalists who refuse to stay in their lane.
The day I realized I was a commodity
I remember a specific interview in late 2018. It was at a fintech startup in Shoreditch called Zego. I was pitching myself as a ‘Growth Specialist.’ I had all the buzzwords. I talked about LTV/CAC ratios and funnel optimization like I was reading from a textbook. About twenty minutes in, the CTO asked me a very basic question about how their specific API would handle a certain type of data call. I blanked. I told him, ‘That’s not really my area, I’m the strategy guy.’
The look on his face was pure pity. I didn’t get the job. I walked out into the rain feeling like a total fraud. I realized that my ‘specialization’ was just a fancy way of saying I was too lazy to understand the actual mechanics of the business. I was a specialist in a vacuum. That’s when I started obsessively learning the boring stuff—basic SQL, how to read a P&L statement, and even some light project management. I stopped trying to be the ‘best’ at growth and started trying to be the person who could talk to the engineers, the accountants, and the customers without needing a translator.
I know people will disagree with this, but I think most junior ‘specialist’ roles should just be eliminated at this point. If you’re 22 and your only skill is writing social media captions, you don’t have a career. You have a hobby that is being automated as we speak. I actively tell my younger friends to avoid ‘niche’ entry-level roles like the plague. They are dead ends.
The T-Shape is about being a bridge, not a pillar
What I mean is—actually, let me put it differently. A specialist is a pillar. It holds up one specific part of the roof. If the ground shifts, the pillar cracks. A T-shaped professional is a bridge. They have a foundation (the vertical bar), but their real value is the horizontal span that connects different parts of the organization.
I tested this recently. I had a project that required a complex data visualization for a client. I contacted a ‘Specialized Data Viz Consultant’ on Upwork who quoted me $1,200 and a five-day turnaround. Then I spent 45 minutes with Claude and a basic understanding of how to structure a CSV file. I’m not a coder. I’m not a data scientist. But I know enough about data structures to prompt correctly and enough about design to know when the output looks like garbage. I did the job in an hour for the cost of my monthly subscription. That specialist didn’t lose the job to AI; they lost it to a generalist who knew how to use AI.
The horizontal bar is your insurance policy.
Speaking of tools, I absolutely hate Notion. Everyone tells me I should use it to ‘organize my life,’ but it’s a bloated, slow mess that feels like it was designed by people who love meetings more than work. I use Apple Notes and a physical notebook because they don’t have a three-second loading lag just to write down a grocery list. I know that’s an irrational hill to die on, but there it is.
Why deep expertise is a dangerous comfort zone
Anyway, back to the point. There is a specific kind of arrogance that comes with being a specialist. You think your ‘depth’ protects you. But in the AI era, depth is the easiest thing to replicate. What’s hard to replicate is context. AI can write a legal brief, but it doesn’t know that the judge in this specific county hates it when lawyers use too many Latin phrases. It doesn’t know that the client is actually terrified of going to trial and just wants a quiet settlement. That’s the horizontal bar—the empathy, the intuition, and the cross-disciplinary knowledge.
I genuinely believe most copywriters under 30 should just quit and go into trade school or learn project management. Writing isn’t a career anymore; it’s a feature of a person. If you can’t do anything besides write, you’re in trouble. You need to be a writer who understands unit economics. Or a writer who can build their own landing pages. The ‘just a writer’ is over.
The market no longer pays for what you know; it pays for how you connect what you know to everything else.
I used to think that being a ‘Jack of all trades’ was a sign of a lack of focus. I was completely wrong. It’s a sign of curiosity. And in a world where the technical skills (the vertical bar) are being commoditized by LLMs, curiosity is the only thing that keeps you relevant. If you aren’t spending at least 20% of your time learning something completely outside your ‘lane,’ you are becoming obsolete.
The ‘Testing’ Phase (Or how I wasted 14 hours)
Last month, I decided to track exactly how I spent my time to see if I was actually living this T-shaped philosophy or just talking about it. I tracked 18 different types of tasks over 21 days. I used a simple spreadsheet—nothing fancy. Here is what I found:
- 40% of my time was spent on my ‘core’ skill (the vertical bar).
- 35% was spent on ‘connective’ work—translating requirements between departments.
- 15% was spent learning tools I had no business using (like basic Framer design).
- 10% was just pure waste.
The 15% spent on ‘useless’ learning actually saved me about $3,000 in freelancer fees that month. It wasn’t that I became an expert in Framer; I just learned enough to not be helpless. That is the secret. You don’t need to be an expert in everything. You just need to be ‘dangerous’ in many things.
I know some people will say this leads to burnout. They’ll say you can’t possibly keep up with everything. And they’re right—you can’t. But you don’t have to keep up with everything. You just have to keep up with the intersections. The magic happens where two unrelated fields touch. If you’re the only person in the room who understands both the marketing budget and the technical debt of the legacy codebase, you are the most important person in that room. You are the one who survives the layoffs.
It’s a brutal way to look at it, but the era of the ‘comfortable specialist’ is over. You can either be a tool that the AI uses, or the person who knows which tools to pick up and how to make them work together. I know which one I’d rather be.
Is this sustainable long-term? I have no idea. Maybe eventually the AI gets good at the ‘connective’ stuff too. But for now, being a messy, curious, T-shaped generalist is the only thing that feels like a solid bet.
Stop trying to be the best in the world at one thing. Start being the only person who can do three things at once. That’s it. That’s the whole trick.