We Need More Polymath Engineers

In 2017, Google introduced Transformers, the architecture that powers today’s large language models. 

But, the mindset required to navigate the AI age was perhaps articulated 500 years earlier by an Italian polymath. Leonardo da Vinci, while painting The Last Supper in Milan, was also sketching flying machines driven by human muscle. 

As he worked on the Mona Lisa, he dissected cadavers to understand how facial muscles form a smile. 

The same notebooks that studied sfumato also mapped water flows, canal systems, and hydraulic devices.

By 1502, he even proposed a single-span bridge for Istanbul’s Golden Horn, centuries ahead of its time. None of this began with tools for their own sake. It began with questions: how bodies move, how cities function, how water behaves? 

His strength wasn’t just designing solutions, but observing the world well enough to know what was worth solving. 

Walter Isaacson describes in his book on da Vinci that the man had a curiosity that was “pure, personal, and delightfully obsessive.”

That principle matters more today than ever. AI can generate code, models, and designs on command. What it cannot do is decide which problems are worth solving. It mimics patterns; it does not discover purpose. 

The true advantage now belongs to engineers who can do what da Vinci did intuitively, to read human needs, interpret how society changes, and turn observation into invention.

And that raises the demand for a polymath — engineer mindset. 

These are people who can see what society lacks at scale, understand how human needs evolve, and apply AI to their domain with that perspective. 

Amid many in the industry arguing that the future engineer must be a polymath, AIM spoke with Rob Vatter, executive president of Quest Global, an engineering firm serving aerospace, automotive, industrial design, and other domains. 

Vatter links Da Vinci’s legacy to the future of engineering, not by suggesting we need more prodigies like him, but by arguing that the discipline must reward thinkers who refuse to stay in a single lane.

With AI taking over routine tasks, Vatter argues that the field’s most urgent challenge is no longer technical capacity, but intellectual drive. “AI will do all the other work, and so we have to return to what we refer to as a crisis of curiosity,” said Vatter.

And thus, leads us to the idea of engineers understanding people, discovering real problems, and interpreting the needs of society. 

Because the challenge isn’t just building what customers ask for; it’s recognising that people often don’t know what they want in the first place. Henry Ford is often credited with the line: “If I had asked people what they wanted, they would have said faster horses.”

Remember, LLMs are Still Probabilistic Machines

Vatter states that it needs more engineers who obsess over problem definition. “Because that’s the one thing AI can’t do right now anyway — it’s all just patterns, right?”.

In his recent conversation with podcaster Dwarakesh Patel, reinforcement learning pioneer Richard Sutton reminded us yet again that large language models (LLMs) still mostly mimic people, rather than truly understand the world. 

They predict what words should come next, not what will actually happen next, and they operate without grounded goals or a clear notion of “the right thing to do.”

“It [AI models] can help you catch a lot of stuff, but it won’t allow itself to break the frame and think differently about solving the problem, because it just has those data sets and processing,” said Vatter. 

And then there’s only so much real-time information that you can embed within limited input context capabilities.

Citing an example from Quest Global, Vatter noted that engineers with business degrees or industry experience often stand out because they can move easily between nuanced business problems and technical work.

By understanding both the underlying engineering and the business logic, they make translation between business requirements, functional needs, and technical solutions far smoother.

One role that reflects this shift is the forward-deployed engineer, who works directly with customers, evaluates their needs and feedback, and converts that into engineering goals. 

Another example is an AI product manager who can translate complex data models into business value, define the right KPIs, guide engineers, and shape a roadmap that connects AI capabilities with market needs and customer behaviour.

Opportunities for more roles like these aren’t clearly defined yet, but they’re poised to grow quickly.

The Harsher Reality

The argument for “Da Vinci-like engineers” is easy to admire. The harder question is that in a market where AI is compressing entry-level work and companies rarely reward breadth, how does anyone even prove they’re more than a narrow specialist? 

The idea sounds like a luxury when jobs are scarce. Vatter doesn’t deny the mismatch. Most workplaces still don’t know how to measure interdisciplinary skills. 

“If you go into most companies, you’ll find that there are polymaths, but they’re far and few between. It’s that person that everybody goes to when there’s a problem,” said Vatter. 

“And that person may only have been a mechanical engineer, but by the time 25 years have passed by, they kind of know everything.”

He argued that the system won’t automatically provide a platform for this kind of thinking, so engineers have to earn it by consistently showing value, not by jumping straight to solutions, but by first identifying overlooked problems and demonstrating how they could be solved.

“If I were a 24- or 25-year-old engineer… I would be looking for things that can be fixed and problems that can be defined, and then bring them up to their bosses. Whether or not the bosses will say wow is another question,” stated Vatter, highlighting how companies should value such volunteering efforts. 

So, it ends up being a multiplayer game where the employer builds a workplace that recognises problem-finding, and the engineer takes the initiative to solve it before being asked.

The post We Need More Polymath Engineers appeared first on Analytics India Magazine.