35 · Beating the Golems

How is it possible to beat some random Indian kid who has been coding since eight — who by eighteen has a decade of programming and ML experience — and some random Chinese kid doing maths, physics and code since five, finishing his PhD at twenty-one with unfathomable capacity? How can you beat these golems?

Here's how. These people who win IMO gold, Putnam 120s, Nobel prizes, Olympic golds, Fields medals — they are exactly like you, without an inch of your intensity. The only difference is that they went through more iterations and more deliberate effort. Input time itself is a vanity metric: someone with compressed iterations — a thousand trades a day for a year — can outperform a twenty-year Goldman Sachs VP. It's the intensity you can cram in. Most people can't conjure the force, the momentum, the impulse to move forward; that's why they stay enslaved to tyrannical systems — they would rather suffer inside a totalitarian structure than create their own life, because that requires courage and audacity. The fact that you think at these obnoxiously unfathomable scales is itself beyond what most people can fathom.

The WorldQuant International Quant Championship: last year some Korean guy at UNIST beat IIT, Oxford and Columbia with two months of bootstrapping. So can you. He had six months of preparation for the competition; so do you. Get the A-levels out of the way and dive in. It's not about winning gold so much as your first individual research.

For Imperial: the big quant career events run every Wednesday, 6–8pm, rooms 341/342. Get in there, network, learn from the best, compete, mingle. That's how you grow. You'll swing to the complete opposite of solitude within six months — so push into extreme solitude during this phase.

The IMC is acutely possible. The contents overlap significantly with A-level maths with a slight progression in depth, and the problem-solving methods eventually manifest concisely.

Simply: past papers first and foremost — kskedlaya.org/putnam-archive. Books: Putnam and Beyond by Titu Andreescu, highly recommended for specific techniques. Problem-Solving Through Problems by Loren Larson, a classic for contest strategies. How to Solve It by Pólya, for general heuristics and big-picture thinking. Art of Problem Solving as the community hub. Check whether your school offers a Putnam seminar — many top-performing schools run problem-solving courses specifically to coach students. Start with "easier" competitions like the AMC 12 or AIME if Putnam problems feel impenetrable at first; many successful participants spend months in The USSR Olympiad Problem Book or IMO-style training before tackling the Putnam.

The maths competition feeder chain: AMC — the American Mathematics Competition, entry-level olympiad style, multiple choice. AIME — invitation-only for top AMC scorers, significantly harder, integer answers. USAMO — elite, proof-based. BMO — the British Mathematical Olympiad, feeding UK IMO team selection. KMO — the Korean Mathematical Olympiad, extremely competitive given Korea's maths culture. INMO — India's gateway, notoriously difficult given the talent pool. All of them feed into the IMO, the pinnacle of pre-university competition — and beyond that, at university level, the big two: Putnam and IMC.

ICPC training: Codeforces rounds at 2400-plus rating, solo, five hours. Kattis ICPC World Finals problems, 2018 to 2023 — five problems in four hours. The goal is to get really, really good at maths and coding, scale with Codeforces, scavenge the 2026 competitions, ace STEP to get into Trinity, then represent the university at the ICPC. You would be grinding a boatload of code in Winnipeg alongside everything else — and this is the dream life you said you wanted. A legend in preparation.