There was a quiet intensity in the room and a universal understanding about how influential the man with the easy smile across the table was. At the center sat Mark Birkhead, a Wake Forest alumnus visiting campus for the first time in over 20 years. You wouldn’t guess he’s a data guy just by looking at him.

I was one of a few students invited to share lunch with Birkhead. To my surprise, the atmosphere was surprisingly casual—light conversation, insights from the data world and some life advice—none of the pressure you’d expect from a high-stakes interview. We kicked things off by chatting about TV shows. Naturally, the latest “White Lotus” episode and its Duke reference came up before we dove into deeper questions about the industry and job market.

Birkhead is the firmwide chief data officer at JPMorgan Chase & Co. (JPMC) and a member of the Wake Forest MBA Class of 2000. Unlike the typical management track, Birkhead’s passion for data began during an internship where he learned to code and solve analytical problems. That experience launched a career that would lead him to build one of Wall Street’s pioneering data teams.

His departure from the conventional business route ultimately earned him recognition as one of the Data Leaders 100 by CDO Magazine—a title that reflects his leadership at a firm processing billions of dollars in transactions every day.

On Thursday, April 3, Birkhead spoke at Wake Forest about JPMC’s integration of artificial intelligence (AI) and machine learning (ML), addressing questions from both students and faculty.

“We have eighteen exabytes of data at JP Morgan—a lot of it collected in just the past three years,” Birkhead said.

An exabyte is 2 to the 60th power. That’s 1,000,000,000,000,000,000 bytes—or a quintillion. Multiply that by 18, and you start to get a sense of the scale.

The event also offered a glimpse into how JPMC applies data and advanced analytics across a wide range of operations, illustrating how financial institutions are positioning themselves at the forefront of the AI revolution.

And from all this, I want to leave you with what I believe are the most important takeaways for students and future early-career professionals.

Coding Should Be Mandatory for Students

Coding should be a mandatory requirement in college, undergraduate or graduate. It’s the elephant in the room, and it’s time students pay attention.

Everyone should consider learning to code—it’s simply where the job market is heading. With AI and machine learning (ML) evolving fast, companies are constantly updating models.

“We have more than 400 AI/ML models at JPMC, and sometimes we need to update them every two to three months,” Birkhead said. “We have about a dozen generative AI models in development.”

Yes, generative AI may replace certain jobs. But learning how to code reduces your risk of being phased out—even if it doesn’t feel immediately relevant.

At Wake Forest, most students start with CSC 111: Introduction to Computer Science. It won’t teach you to build or differentiate large language models like ChatGPT or Copilot, but it does give you foundational exposure. And that matters.

Skipping the basics is like learning English through rap lyrics or doing calculus without knowing arithmetic. You have to start with data types, loops, and functions — core concepts that make up the language of code.

These introductory classes use Java, a structured language that is tough at first but teaches discipline and clean coding. It’s like your high school English teacher insisting on full paragraphs instead of bullet points. There’s a reason for it.

I believe Python should be taught alongside Java, not as a replacement but as a complement. While Java builds structure, Python is what many companies actually use. It’s readable, job-ready, and connects easily to advanced libraries and AI tools—many of which are free and open-sourced by tech giants like Google and IBM.

The cherry on top? You don’t have to go it alone. Tools like ChatGPT can help you learn and use these packages—even the ones that power the very AI models behind your screen.

There’s a Good Argument to Learn Coding for a Job

JPMC even hosts biannual hackathons, inviting analytically-minded students to solve real-world problems. These events often lead to internships, rotations and full-time offers.

You might think coding isn’t relevant to your dream finance job on Wall Street. You’re busy mastering Excel and memorizing the Green Book of 400 interview questions. And sure, ChatGPT still struggles with simple math—so what’s the rush?

But Birkhead warns the shift is coming. We’re in the early phase of AI adoption, but as companies develop specialized tools trained on proprietary data, the line between analyst and algorithm will blur.

JPMC is already investing in retrieval-augmented generation (RAG) models and small language models trained for specific tasks. Will they outdo my calculus skills? I’m betting yes—sooner rather than later.

The good news: learning to code isn’t that hard. If you’ve ever written an academic paper, you’ve already put in more time than it takes to grasp the basics of Python.

Practical tip—check out LinkedIn Learning, free with your Wake Forest account. Take a course, earn a certificate and add it to your resume. It’s a smart way to future-proof your skillset.

You want to be kept in the loop

A recent paper by researchers from the University of California, San Diego—currently under review—reported that GPT-4.5 has passed a modern version of the Turing Test. Its findings once again challenge our assumptions about how far the AI revolution can go—and what remains of uniquely human intelligence and labor.

“Some jobs in customer service and call centers may be gone because these are repetitive tasks that Gen AI can tackle,” Birkhead said. “However, we hope people will move into positions where human involvement is necessary in the loop.”

“We need people overseeing these models,” Birkhead added, “because we need to be very confident that these models give accurate and dependable feedback.”

The demand for skilled analytical professionals—those who understand computer languages and can translate outputs into actionable business solutions—will only increase. Now is the time when learning formulas in Excel is just knocking on the door, while learning to code and explain the results has become the meat and bones of an entry-level analyst role.

If you start now, it’s not too late to embrace the world of coding. And there’s always a chance that a top firm like JPMC will be waiting for you—not for the Excel monkey, but for the flexible coder.