Action at a Distance (Physics)

In the AI Story

The doctrine's persistence from Newton's Principia (1687) through the mid-19th century reflected its stunning predictive success. Lagrangian and Hamiltonian mechanics, built on action-at-a-distance principles, could calculate planetary orbits, predict eclipses, and describe mechanical systems with such precision that the philosophical unease about mechanism seemed like mere metaphysical fastidiousness. Practical scientists could ignore the how and focus on the what—measuring forces, calculating trajectories, predicting outcomes—without ever confronting the incoherence of instantaneous influence across empty space. The mathematics worked; the philosophy was optional. This is exactly the stance contemporary AI analysis adopts: measure productivity, calculate displacement, benchmark capability, and treat the how (what happens in the creative space between builder and system) as operationally irrelevant.

Newton's own discomfort is documented in his correspondence with Richard Bentley, chaplain to the Bishop of Worcester, who was defending Newtonian physics against theological objections. Newton wrote: 'That one body may act upon another at a distance through a vacuum without the mediation of anything else...is to me so great an absurdity that I believe no man who has in philosophical matters a competent faculty of thinking can ever fall into it.' He proposed the ether—a subtle, invisible medium filling space—as a possible mediator, but could not specify its properties or incorporate it into his mathematical framework. The contradiction persisted: the theory's founder rejected its central premise even while defending the theory's empirical adequacy. Later natural philosophers mostly ignored the discomfort, treating action-at-a-distance as a regrettable mystery that successful prediction made tolerable.

Faraday's achievement was insisting that philosophical coherence and empirical adequacy were both required—that a framework giving correct numerical results while postulating incoherent mechanisms was not a successful theory but an incomplete one awaiting deeper understanding. His iron filings demonstrations were not refutations of Coulomb's law (the numerical predictions remained valid) but revelations of what the law described through: not empty space but a field whose structure the law's mathematical form encoded without acknowledging. Maxwell's synthesis preserved all the successful predictions while replacing the absurd mechanism with a coherent one. The forces between charges are real, but they are mediated—transmitted through continuous field, not mysteriously leaping across voids. Action-at-a-distance was not wrong about the forces; it was wrong about the space.

The AI parallel is structural: current frameworks describe forces (productivity gains, job displacement, capability expansion) between poles (human workers, AI systems) and treat the space between as empty conduit. The measurements are often accurate—developers do produce more code, lawyers do draft briefs faster, customer service throughput does increase. But the framework is radically incomplete because it describes outcomes while ignoring the medium through which they are produced. The builder's experience of working with AI—the creative tensions, the momentum, the identity transformation—are dismissed as subjective epiphenomena, interesting for psychology perhaps but irrelevant to understanding the technological transition itself. This is the same error that dismissed the iron filings as pedagogical decoration. The patterns are evidence of field structure, and the field is where the transformation is happening. You cannot understand the transformation by measuring the poles and ignoring the medium between them any more than you could understand electromagnetism by calculating forces while denying that fields exist.

Origin

The framework has roots in ancient atomism (Democritus, Lucretius) but achieved mathematical formulation with Newton's inverse-square law of gravitation (1687). Coulomb extended it to electrostatics (1785), establishing that electrical forces also followed inverse-square laws. The 18th-century elaboration—Laplace's potential theory, Lagrange's mechanics—mathematized the doctrine without addressing its philosophical foundation. Only in the 1830s-40s, as Faraday's experimental work accumulated, did the framework face serious empirical challenge. Even then, the Continental mathematical tradition (Ampère, Wilhelm Weber, Franz Neumann) resisted field theory for decades, preferring to develop increasingly complex action-at-a-distance formulas rather than adopt Faraday's conceptual revolution. The framework's death was gradual: Maxwell's equations (1865) provided the mathematical alternative, Hertz's 1887 demonstration of electromagnetic waves confirmed field-propagation at finite speed (incompatible with instantaneous action-at-a-distance), and by 1900 field theory had triumphed. Quantum mechanics and general relativity completed the inversion: all fundamental interactions are now understood as field phenomena, with action-at-a-distance surviving only as a classical approximation valid under limited conditions.

Key Ideas

Predictive success does not guarantee conceptual adequacy. A theory can give correct numerical results while resting on incoherent premises—warning that current AI productivity measurements may be empirically accurate yet theoretically blind to the actual mechanism of transformation.

Mathematical sophistication can conceal physical absurdity. Elegant formulas may encode relationships that make no physical sense when their implications are examined—suggesting that sophisticated AI benchmarks and economic models may obscure rather than clarify what is actually happening in creative collaboration.

Mechanism matters, not just prediction. Understanding how an effect is produced is categorically different from calculating that it is produced—establishing that AI-transition analysis must investigate the creative process (field dynamics) rather than merely measuring outcomes (pole displacements).

Empty space as default assumption. The void is what you assume when you are not looking carefully—applied to AI, this means the 'empty interface' assumption (that the space between user and system is neutral infrastructure) is likely a failure of investigation rather than an accurate description.

Philosophical coherence as scientific necessity. Faraday's insistence that absurd premises disqualify otherwise successful theories sets a standard: AI frameworks postulating human-machine interaction without investigating the interaction's structure are philosophically inadequate regardless of their predictive accuracy.