There's an old joke in finance about "poets and quants" — the idea that investment banking required both the analytical rigor of quantitative analysts and the storytelling ability of people who could explain it to clients.
The best professionals could do both. They were rare.
I keep coming back to this framing as I think about the skillsets that will matter most in the agentic AI era.
The New Poet-Quant Divide
In the agentic era, the divide isn't between quantitative and humanistic skills in the traditional sense. It's between two capabilities that most educational and professional development systems treat as separate.
The ability to understand systems. How does the technology work? What are its real capabilities and limitations? How do components interact? What does the data model look like? What can you actually build, and what's harder than it looks? This is the quant side — not pure math, but structural, systems-oriented thinking.
The ability to understand humans and organizations. What do people actually need, and why? How does an organization change? What's the real stakeholder concern underneath the stated objection? What does adoption actually require? This is the poet side — not pure literary sensibility, but human and organizational understanding.
In the AI era, both are necessary because AI has fundamentally shifted what human work is.
Why Both Matter More Now
AI is handling more of the routine cognitive work — research, drafting, synthesis, pattern recognition. What remains for humans is the work that requires judgment at the intersection of systems and humans.
You need to understand the technical systems well enough to know what AI can and can't do — to direct it toward the right problems, to evaluate its outputs, to know when it's wrong.
And you need to understand the human and organizational systems well enough to know what actually needs to happen — what the customer really needs, what the stakeholder is actually worried about, what change the organization will and won't accept.
Pure technologists who don't understand organizational dynamics will build systems that don't get adopted. Pure humanists who don't understand the technical systems will be directing AI toward the wrong problems. The combination is where the value is.
How to Develop Both
The good news: these are learnable skills, and the learning paths are available.
For the systems side: get your hands dirty with AI tools. Build things. Understand the integration architectures your organization uses. You don't need to be a deep technical expert — you need practical intuition about what systems can do.
For the human/organizational side: the skills are built through attention and practice. Who actually influences decisions in this organization? What does adoption of previous technology look like? What do people do when they're uncomfortable with change?
The goal is fluency in both worlds — not expertise in either. Enough understanding of the technical to evaluate what's possible. Enough understanding of the human to evaluate what's needed.
That combination is the new poet-quant. And it is rare.
Views are personal.