The Lost Einsteins: How Banking's Obsession With the Past Is Driving Away the Talent That Will Define Its Future

(Disclaimer: long post but a topic I believe is worth thinking about)

Clay Christensen famously warned in The Innovator's Dilemma that even the most successful companies fall apart when they can't escape the structures that once made them great. His argument wasn't about talent or resources — it was about architecture. Leading firms lose to scrappier competitors not because they're outgunned, but because their organizational DNA is optimized for a world that no longer exists.

I've spent the last 4.5 years across three investment banks — micro boutique, boutique, and bulge bracket — working across Tech, SaaS, and Retail & Consumer coverage. Each firm had its own culture, deal flow, and resource base. But one thing was consistent across all three: no clear direction on how to actually use AI in deal team workstreams.

That gap bothered me more the longer I sat with it.

I've followed the development of generative AI closely since ChatGPT's launch in late 2022. What started as a novelty has compounded into something that touches virtually every professional workflow — research, writing, analysis, synthesis. The refinement over the last two and a half years alone has been staggering.

And the market has noticed. We've watched Nvidia earnings calls single-handedly set the tone for weeks of trading. Forward guidance on AI infrastructure spending doesn't just move one stock — it moves the tape. The market's read is clear: AI will be the dominant economic force for the next several decades.

So why are the institutions that live and breathe capital allocation so slow to deploy these tools internally?

To be fair — and I know the WSO crowd will call me out if I'm painting with too broad a brush — some firms are further along than others. Several bulge brackets have proprietary AI initiatives they're keeping close to the chest for obvious competitive reasons. That's rational. But for an industry with a near-pathological obsession with precision and efficiency, the lag in AI adoption at the deal team level is hard to explain away.

The slow rollout is only half the problem, though. The deeper issue is structural.

The organizational hierarchy most firms still operate under was engineered for a different era entirely — the industrial mid-1800s, when competitive advantage meant coordinating large groups of people to execute repeatable tasks reliably. Command-and-control worked when scale was the moat. But that logic is being dismantled. AI doesn't require specialized training to deploy, which means innovation is no longer the exclusive property of leadership — it can emerge from anywhere in the organization. And as AI absorbs more of the routine and technical workload, the capabilities that can't be automated are moving to the center: creative thinking, entrepreneurial instinct, curiosity, the ability to communicate and connect ideas across contexts. The irony is that we've spent decades calling these "soft" skills — as if they were secondary. In the years ahead, they'll be some of the hardest to find and the most expensive to lose. There's a term I've become obsessed with that captures exactly who gets left behind when organizations fail to make this shift: The Lost Einsteins.

Which brings me to hiring — because the structural problem and the talent problem are the same problem wearing different clothes. Banking has long fetishized a very specific type of intelligence — the kind that can mentally estimate the number of windows across every office building in Manhattan. Those questions were never really about windows. They were proxies for raw analytical horsepower and composure under pressure — reasonable screens for a world where the edge came from processing information faster than the next person. But that world is fading fast. In an AI-driven environment, the cognitive tasks those brainteasers were designed to identify can increasingly be offloaded to a model in seconds. What can't be replicated is the entrepreneurial mindset — the ability to ask better questions, spot non-obvious angles, and build something meaningful from ambiguity. Firms still optimizing their hiring for yesterday's skill set will find themselves stocked with brilliant people who are perfectly equipped for workflows being automated underneath them. And retaining the talent that actually thrives in this new environment gets harder by the day — because the people with the creative, adaptive instincts AI demands will eventually stop waiting around for organizations that don't know how to use them. Those are your Lost Einsteins. And once they're gone, they're not coming back.

If you're trying to break in — as a student or career pivoter — this is worth internalizing before your next interview. The definition of top banking talent is evolving, and it's happening faster than most recruiting cycles acknowledge.

Opening the floor for discussion – challenging thoughts are encouraged. Would also be curious to hear from anyone currently working at a firm that has started to shift more toward this perspective.