The Preschool Arms Race and the Algorithm of Ambition
Imagine two search algorithms competing to find the treasure in a vast, foggy landscape. Breadth-First Search (BFS) sends out hundreds of drones to scan every inch of the surface, methodically mapping the obvious. Depth-First Search (DFS) sends a single probe drilling straight down, risking total failure for the chance of striking a hidden vein of gold.
Now imagine the landscape is the state space of possible human innovation, and the drones are children. Tiger moms, hyper invested parents optimizing their offspring's trajectories, are the programmers. Their choice? Deploy BFS. Always BFS.
Why? Because in a world where parents correlate their strategies (i.e., all push kids toward the same SAT-optimized, resume-padding, YC class, 'safe' milestones), BFS dominates. This creates a pseudo-tragedy of the commons: while individual parents rationally chase visible, low-risk gains, the collective outcome is a world starved of the deep, weird, uncorrelated innovations that require DFS solutions.
BFS Parenting: How Tiger Moms Colonize the Local Maxima
The BFS approach to child-rearing is recognizable to anyone who's seen a kindergarten resume boasting Mandarin, violin, and quantum physics (taught via nice child friendly books):
- Checkbox Childhoods — they should master all high-status, testable skills (math, coding, classical instruments)
- Risk Aversion — Avoid 'wasting time' on niche interests (competitive skateboarding, historical whaling trivia)
- Orthogonal Bragworthiness — Ensure activities are diverse enough to signal "well-roundedness" but conventional enough to avoid raising eyebrows
This strategy isn't irrational. BFS works if your goal is to maximise admission to MIT or entry into quant/tech. It optimizes for exploiting known peaks (existing prestige hierarchies) rather than exploring unseen peaks (new fields).
The result? A generation of kids climbing the same hills, slightly faster, slightly more impressive, while the landscape's true mountains remain undeveloped, distant, invisible, go untouched. A generation of kids vying to drop out of Stanford and get VC funding for a GPT wrapper, FTX, and leet code grind.
DFS and the Economics of Weirdness
Depth-First Search is the domain of the obsessive, the eccentric, the uncorrelated. It's the 12-year-old who spends 6 hours a day writing fan fiction, reading LessWrong, the teen who converts their basement into a CRISPR lab because he thinks bioweapons are cool, the autodidact mastering obscure Liszt etudes instead of exam-friendly Chopin.
DFS is high-risk, high-reward. There is individual risk in sinking time into a dead end (RIP, my unfinished fanfiction at 17, lol), but occasional societal rewards: Taleb's Black Swan innovation (Turing's obsession with computable numbers).
Tiger moms disincentivize DFS. Why?
- Correlated payoffs — when all parents optimize for Ivy League admissions, they create a Nash equilibrium where deviating (DFS) is punished. A kid studying AI art in 2010 was a visionary, in 2023 it's another checkbox.
- Signalling costs — DFS projects are illegible to admissions committees until after they succeed. For every "started a successful app" teen, there are 100 who just "wasted time on computers". MIT Maker Portfolios, to me, are the best example of DFS or more than talented BFS strategies.
- Privilege tax — DFS becomes a luxury only the privileged (or neglectful) can afford.
The Innovation Desert: When All Peaks Are Overcrowded
In Seeing Like a State, James Scott warns of monocultures replacing diverse, local ecosystems. Similarly, BFS parenting creates an innovation monoculture.
Everyone learns Python. Result? A million competent data analysts. No one learns Haskell, so the next groundbreaking functional programming paradigm goes undiscovered.
All musicians play piano/violin. No one masters the theremin, so electronic music evolves slower, most music will comprise of piano and violins.
Even worse, BFS creates hyper-competition on crowded peaks. Kids grind LeetCode to out-code peers, but 90% of their effort is spent one-upping rivals, not expanding the frontier. It's Red Queen innovation: running faster to stay in the same place.
Why They're Half Right...
"BUT BFS BUILDS FOUNDATIONAL SKILLS!!" — True. Breadth matters. The Wright Brothers leveraged physics, engineering, and bicycle mechanics. However their DFS-like obsession with flight (ditching safer careers) was key.
"MOST DFS FAILS!!" — Also True. However societal progress relies on portfolio risk. Taleb talks about systems, markets, societies needing to be antifragile at the cost of cells, companies, people being fragile. If 1000 DFS kids 'fail' into niche expertise, their pooled knowledge becomes new fields, new experts, or politicians who are actually good.
"TIGER MOMS PRODUCED ASIAN TECH MIRACLES!!" — Yes, but these are BFS miracles from playing catch up (incremental optimization, not paradigm shifts). China dominates AI applications but, as of right now, relies on algorithms mostly invented elsewhere.
Cultivating a Garden of Forking Paths
How to encourage DFS without abandoning BFS?
Reward Weirdness — Colleges and employers could value narratives of obsession over checklists. Did you spend 3 years building a Destiny 2 discord democracy? That's a governance simulation.
Delayed Specialisation — More similar to Finland: broad curricula early, deep dives later.
Parental Bravery — Let kids 'waste time' post-metric. Minecraft modding taught a generation to code.
The DFS Beneath BFS
The next Einstein is not grinding leet code problem sets, they're daydreaming on a park bench around their friends, chasing a thread only they will see. Tiger moms are not evil, they're trapped in a coordination problem. By marginally valuing DFS, we might yet escape the local maxima.