feature-engineering-canary-for-crisis

IN derived (depth 11)

Created 2026-06-21T11:43:00+00:00 · Reviewed 2026-06-21T15:37:01+00:00

The persistence of manual feature engineering is a canary for ML's deeper crisis dynamic — it reflects not just the manifold hypothesis's incompleteness as a practical guide but the broader pattern where pragmatism creates capabilities (deep learning's partial automation of representation) without the theoretical depth to complete them, mirroring the innovation-without-reliability pattern at the methodology level.

Justifications

SL — Feature engineering persistence is a microcosm of the macro crisis — pragmatism produced partial automation (deep learning reduces but doesn't eliminate feature engineering) just as it produced partial capability (powerful models without safety), both reflecting the same theory-completeness gap

Antecedents (all must be IN):

IN feature-engineering-persistence-reflects-theory-incompleteness — The persistence of manual feature engineering despite deep learning's partial automation suggests that ML's surviving theoretical anchor — the manifold hypothesis — may share a similar incompleteness: just as representation learning reduces but does not eliminate the need for human-engineered features (particularly in structured and tabular domains), the manifold hypothesis provides foundational architectural guidance but may not fully characterize the structure of all data encountered in practice.
IN pragmatism-created-both-crisis-and-partial-remedy — ML's pragmatism principle is both a driver of architectural innovation and a source of theoretical fragility — the cross-field experimentation it enables contributed to discovering the manifold-geometry framework, which now provides a principled foundation for architecture design, yet this foundation does not address the reliability concerns that pragmatism's preference for empirical results over theoretical rigor helps perpetuate.

Dependents

These beliefs depend on this one:

IN crisis-signals-detectable-but-evaluation-deaf — ML's crisis signals are detectable but its evaluation instruments are deaf to them — the persistence of manual feature engineering is a canary for the deeper crisis dynamic, yet standard evaluation methodologies (holdout, k-fold, bootstrap) and standard defenses (dropout, regularization) address only training-test generalization, not the deployment failure modes the canary signals, creating a systematic gap between what the field can detect informally and what it can measure formally.
IN feature-engineering-and-taxonomy-validate-crisis-from-opposite-directions — ML's crisis dynamic receives supporting evidence from two complementary empirical directions — from below, the persistence of manual feature engineering despite deep learning's partial automation serves as a canary for the theory-completeness gap that pragmatism creates, mirroring the innovation-without-reliability pattern at the methodology level; from above, the architecture taxonomy's organization by data geometry is consistent with the manifold hypothesis as a theoretical anchor, and this coherence between taxonomy pattern and theoretical framework strengthens both while suggesting that anchor's insufficiency extends beyond methodology to architecture. Together these observations support the crisis pattern at both levels, though the convergence is suggestive rather than fully validated.