Where Wall Street's Best Brains Disappeared To

Over the past two years, the number of times I've been laughed into silence in an interview room has surpassed the total from the previous decade combined.

Last month I sat across from a freshly minted Ivy League grad. Line three of his résumé read "Proficient in Python." I flipped to a random internship report he'd attached, and asked which package he'd used for data cleaning. He paused, then cracked a smile, perfect white teeth, the unshakable confidence of someone who'd never been told "no", and shot back in a breezy, "you're so outdated, professor" tone:

"To be honest, I do all my coding with AI now. I don't really know which specific package it calls. Which AI does your firm use?"

I stared at his full cheeks of collagen and his pristine, market‑untested self‑assurance, and one thought crystallized: If you don't even know what's under the hood, why would I pay you a six‑figure base? I could just ask IT for a ChatGPT Enterprise seat and save enough to cover my year‑end bonus.

Don't get me wrong. I'm not anti‑AI. I use it myself. Pure‑hand coders and prompt‑jockeys both have a place on my team, as long as the work gets done cleanly. You could write C++ in Notepad for all I care. But the real issue boils down to two characters: integrity.

You write "Proficient" – turns out the AI is proficient. You write "Analytical" – turns out the AI did the analysis. How is that any different from buying a paid internship certificate? Oh wait, I forgot: you did buy that too.

This theater of the absurd keeps repeating, and it's forced me to confront two questions. First, is the financial market actually fair? And second, where did all the genuinely sharp people go?

I. Fair? That's Only "After You Place Your Bet"

Plenty of people bash finance as corrupt; plenty others praise it as fair. The fairness crowd usually argues that the market only cares about money, not connections.

As someone who deals with money every day, let me pour some cold water: the financial market was never fair – it's only "ruthlessly symmetric."

What does symmetric mean? You bet right, add to your winners – the market doesn't ask who your father is. You bet wrong, blow up your leverage – the market doesn't care how loud you cry. Liquidation doesn't wait one second. That brutal symmetry is precisely what used to attract the smartest minds.

But when it comes to getting in the door, fairness hasn't been in the same zip code.

Some commenters love to fantasize about "old money and insider resources," assuming we BB types are naive about connections. Let me be blunt: if you have to prove your resources through an interview, they're not real resources. Those semi‑hereditary seats are filled before the job posting ever goes live. The roles you see on our website were always meant for the rest of the world – people who compete on degrees, internships, and certifications, only to realize at the finish line that even their "Proficient in Python" was ghost‑written by an LLM. What else is left to compete on?

So please, don't project your trading losses onto the entire industry. Finance is a legitimate modern service business. It has its grimy corners, sure, but most of the time it's about risk, compliance, and reputation.

For the past three decades, why did top‑tier talent still fight to get in? Leverage. Pure and simple. You didn't need to understand the technology – just how to allocate capital. With 20‑ or 50‑times leverage, cognitive arbitrage printed money faster than a central bank.

That logic, however, is now broken.

II. The Great Migration: Where Did the Sharpest Brains Go?

It pains me to admit this, but I have to: the true cognitive elite has been uprooted, lock, stock and barrel, by AI and hard tech.

Why?

Because finance is fundamentally "second‑hand leverage." It creates no atoms; it only shuffles them. AI and frontier tech, by contrast, are "first‑hand leverage" – they build things from zero to one.

In the old days, the math and physics PhDs from Tsinghua, Peking, the Ivies – they had few choices. Graduation meant either writing models on Wall Street or drafting prospectuses in Lujiazui. Two paths, and the finance side was still the wider one.

Now? On one hand, Citadel and Jane Street are throwing monster packages at quants. On the other, OpenAI, DeepMind, and every big‑tech AI Lab are waving equity grants. How does a 22‑year‑old math prodigy choose?

Just do the arithmetic:
To make Managing Director in finance, the average is 15 years. During that time, you endure compliance reviews, endless pitch‑deck reworks, and client dinners that leave you puking blood. At a top quant firm or AI shop, you can vest equity and achieve financial independence by 30 – no forced smiles, no explaining why you're revising a prospectus at 3 a.m.

The sharpest minds can run those numbers. It's not that high‑cognition people inevitably gravitate to finance; it's that they gravitate to wherever the cognitive leverage ratio is highest. Yesterday it was here. Today it's over there.

Among the "survivors" we manage to hire, only 10% are truly combat‑ready. And most of that 10% joined not because they love finance, but because they missed the AI wave – or because their risk appetite is pathologically low.

The other 90% fall into two categories.

III. Being Green Is Fine. Being Green and a Fraud Is Not.

The recruitment market today is a 90‑10 split.

Strip away the top decile; the remaining nine‑tenths are not just incompetent – they're arrogant about it.

I once had a candidate who couldn't stop bragging about "making five points on options yesterday." Then I asked him to join two tables with 100 million rows each in Excel. He pulled out VLOOKUP and started grinding – his laptop froze, and he blamed IT for giving him a crappy machine. I said, "I thought you were proficient in Office?" He shot back, "VLOOKUP is obsolete – who doesn't use Power Query these days?"

Then why didn't you use it? Silence.

This isn't skill degradation. This is logic amputation.

What's more sickening is the collapse of integrity.

Among the candidates who make it past HR's initial screen – so already a filtered pool – 10% to 30% can't even meet the basic bar of honesty.

• Internships on the résumé? 50,000 RMB a pop, list price; just search "paid internship" on Xiaohongshu, full service available.

• Degrees? Harvard/Yale/Princeton/Stanford/MIT – most expensive, 10 million RMB each; regular Ivy undergrad, 1‑3 million; cheaper UK/HK master's, bundled discounts.

• Interview cheating? AI‑assisted answers are entry‑level. Advanced setups use micro‑earpieces with a remote assistant feeding real‑time search results.

Every year when I visit campuses, I repeat one line: "Don't try to fool me on your résumé. Assume that everything you list, except maybe sports and hobbies, I know, I've done, and I can drill you on until you crack."

That's not cruelty. That's an integrity test. If you're willing to fake "Proficient in Python," would I trust you to validate a public company's financials? Would I let you near a risk model? You'd hand me the next Enron inside a quarter, then shrug and say, "But the AI told me to fill it that way."

IV. I Give Only One Test in My Interviews

Some people say we BB firms only hire PowerPoint weavers and button‑pushers.

Spare me. Day‑trading your own account doesn't make you "in finance"; it makes you a retail investor. Finance is intermediation, allocation, risk control, compliance – a vast collaborative machine.

We don't care how many certificates you've stacked or how many formulas you've memorized. My entire interview now consists of one practical exercise:

A CSV file, 5 million rows, riddled with logical traps, missing values, and counter‑intuitive anomalies. No tool restrictions, AI allowed. One hour. Tell me: is this data usable? If not, what's wrong with it?

In that hour, I've seen two archetypes.

The first: the true high‑cognition operator.
He might not write a single line of code. He opens the CSV, scans the column headers, samples the first 100 rows, then looks up and asks: "What's the business definition for this field? Why does this column have negative values here? What timezone standard is this timestamp using?" He doesn't know the tech, but he knows logic. He decomposes, traces back, questions assumptions.

The second: the AI prompt‑jockey.
He moves fast. Pastes the entire 5‑million‑row file into an AI chat window. Types: "Analyze this CSV for me." The AI errors out; memory limit exceeded. He panics. Turns to me: "Is this file corrupted? Is your system broken?"

See the difference?

Tools are hands and feet. The brain is the only weapon.

Final Word: No Golden Age, Only a Survivor's Game

Finance's golden age is indeed over. The BBs have lost the talent war to tech giants – that's industrial evolution.

But I've never believed this industry is done for. The jobs that lived on information arbitrage are disappearing, but the core abilities – steering risk, spotting fraud, allocating capital – will always be scarce.

To the young guns grinding to break in, I have only one piece of advice:

Ditch the clever tricks.

Don't use AI as a stunt double. Don't treat purchased internships as genuine experience. Don't think you can skate by on the word "Proficient." In an industry that despises risk above all else, "reliable" and "honest" are your last and only premium.

And to the 10% who've already made it through – welcome to the real finance world.

There's no easy money here. There is a real cognitive contest. You don't need to hand‑roll a Transformer from scratch. But you do need to be accountable for every single word on your résumé.

Because the market never shortchanges the truly clear‑headed.

But it will never, ever let a prompt‑jockey walk away unscathed.