The debate centers on the distinction between theoretical complexity and practical execution, specifically regarding 'galactic' algorithms. These are algorithms that possess superior Big-O complexity (the asymptotic growth rate) but are rendered unusable by constant factors so large they exceed the number of atoms in the observable universe. The Exponential Time Hypothesis (ETH) provides a theoretical framework for understanding these limits, suggesting that certain problems require exponential time to solve. The core of the controversy lies in whether these theoretical bounds remain meaningful when the algorithms required to achieve them are physically impossible to implement.
In a recent critique, a math professor argued against what he termed the 'Big-O Cabal,' a perceived culture in computer science that prioritizes asymptotic optimality over practical utility. The author specifically pointed to a recent proof involving OpenAI, noting that the algorithm used in the proof performs a brute-force search for all instances below a 'super-astronomic' threshold. This observation suggests a disconnect where a proof might claim a certain complexity class while relying on a brute-force approach for any input size that a human or current machine could actually process. This 'buries the lede' of the proof, as the theoretical efficiency only kicks in at scales that are effectively unreachable.
Conversely, many in the computer science community argue that this critique misses the point of complexity theory. Commenters on Hacker News noted that Big-O complexity is, by definition, independent of constant thresholds; an algorithm's asymptotic behavior remains the same whether the switchover point to a more efficient method is at ten or at a trillion. They argued that 'unrealistic' math is a standard and valid pursuit in complexity theory, much like pure mathematics. The trade-off remains a fundamental one in the field: whether the goal is to find the absolute mathematical floor of a problem's complexity or to develop algorithms that can actually be executed on physical hardware. While the author views the reliance on galactic constants as a 'losing the plot' moment, others maintain that the pursuit of theoretical purity is a necessary and valid scientific endeavor.