The Best Modeling Tool I Didn’t Learn in Academia

How GenAI changed the way I approach feature engineering, forecasting, and iteration.

Real-world modeling differs starkly from academic code. Image generated with Leonardo AI

In academia, we’re taught to revere clean models, strict workflows, and the slow accumulation of rigor. That discipline has served me well — until I stepped into the world of messy sustainability data and business timelines.

Here, the perfect model is the one that’s good enough to ship. And the best assistant I’ve found for building those models? GenAI.

To be fair, I’d worked with GenAI during academia; but I rigorously tested my model for a year (a year!) before I and my PhD thesis advisor submitted the code to a scientific journal.

GenAI is not awesome because it replaces my thinking—it doesn’t—but because it reflects my thinking back at me. It works faster than me, more playfully, and with surprising insight.

And when it fails, it fails faster than I ever could as a mortal human. GenAI is not just a tool. It’s a mirror.

What Academia Taught Me — And What It Missed

In academia, I was trained to respect structure: define your problem, lay out your assumptions, choose your tools carefully, and build upward — rigor first, clarity second, iteration third.

That works beautifully when you’re building models for theoretical systems. But out here, in the chaotic realm of sustainability and finance, the world doesn’t hand you clean data or clearly defined variables.

You’re often trying to model systems you barely understand — social dynamics, climate vulnerabilities, fragmented balance sheets — all under time pressure, with incomplete information.

I had the tools to formalize. But I needed something to catalyze.

How GenAI Became My Co-Thinker

At first, I used GenAI for small things: rewriting Python docstrings, explaining why a model threw an error, helping me brainstorm variable names.

But slowly, it crept into my upstream workflow — especially during those “I don’t even know where to start” moments.

Here’s what it helped with:

• Feature ideation: I’d describe the problem and dataset, and it would return ten potential features — some obvious, some genuinely novel.

• Scenario framing: I could ask, “What variables might shift under X policy change?” and get useful thought starters for modeling.

• Forecast scaffolding: I’d describe a desired output, and it would sketch out how to structure the model pipeline, which techniques to consider, and what tradeoffs might show up.

Was any of it plug-and-play? Never. But it was always enough to move me forward.

Sometimes GenAI doesn’t solve the problem — it just reminds you there is a solution to the problem.

The Tool That Reflects Your Thinking

What surprised me most is how using GenAI became a mirror.

When I asked it to generate feature sets, I started noticing my own blind spots. When it gave me a model outline, I noticed what it assumed about the data — and whether I agreed.

The conversation itself became part of the modeling process. Not an interruption, not a shortcut — but an extension of my own thinking.

It wasn’t doing the modeling for me. It was helping me see my modeling logic more clearly.

Where It Breaks — and Why That’s Useful

There are things GenAI still can’t do well:

• It fumbles with advanced math, especially dynamic or probabilistic reasoning.

• It can’t prioritize well when information is ambiguous.

• It occasionally invents logic that sounds good but breaks when implemented.

But these flaws are helpful in my opinion.

They highlight what requires human judgment. They nudge me to double-check things I might otherwise gloss over. They make me more cautious in the right places.

From Clean to Curious

In academia, I aimed for elegance. Now I aim for curiosity.

Where I am now, models don’t need to be clean first — they need to be useful first. Especially in sustainability, where uncertainty is high, stakes are real, and the story matters as much as the prediction.

Using GenAI lets me start messy and refine later. It gives me permission to sketch, to brainstorm, and to reshape ideas in motion.

GenAI shifted my modeling from “prove this works” to “explore what’s possible.”

The Bottom Line: Start With the Question, Not the Framework

The best tool I never learned in school isn’t a statistical method.

It’s an approach: collaborative, iterative, and more interested in direction than precision. Except you don’t even need colleagues for this.

GenAI won’t replace modeling discipline. But it does reshape how I use that discipline — and where I invest my energy.

When your goal is clarity, not control, GenAI becomes less of a shortcut — and more of a co-pilot.

---

Wangari’s Curated Reads

• What Spain's power outage tells us about the future of clean energy, knowns

  David Carlin
  . This analysis of Spain's recent blackout offers a vital insight: the clean energy transition isn't just about more renewables—it’s about smarter grids. As Spain rapidly scaled up solar and wind, it neglected to modernize the grid infrastructure that supports them, leaving the system vulnerable. It’s a sharp reminder that building a sustainable future requires attention to the unglamorous backbone of energy systems—resilient, interconnected, and up-to-date infrastructure that can handle both climate shocks and renewable variability.

• Guilty fact: I love chocolate. But I’m less inclined to eat more, after discovering the bittersweet truth, revealed by

  Willow Lynn
  and
  Brightest
  . From child labor and farmer poverty to deforestation and biodiversity loss, it really isn’t pretty. For Wangari Digest readers, it's a compelling case for choosing ethically sourced chocolate and demanding transparency from brands—because even small decisions can push the industry toward justice and sustainability.

• Zen Peacemakers
  have another take on sustainability. The Power of Community. The Gift of Sustainability is a brief but heartfelt message that sustainability isn’t only about the environment—it’s also about sustaining communities rooted in care, practice, and shared purpose. For Wangari Digest readers, it’s a quiet nudge toward the kind of mindful giving that nourishes the deeper roots of sustainability: connection, service, and mutual support.

---

One Smart Question