Roger Martin’s “knowledge funnel” is a very useful model for understanding the human pursuit of knowledge. Man contemplates a new mystery using intuition to infer causality, by trial and error arrives at an inexact approach that somehow seems to tame the new mystery before finally framing the new phenomenon with the objective precision of a rigorous formula. This is the process by which the vast unknown is distilled into bits of knowledge that our puny minds can manipulate.

Putting this process into perspective, one is tempted to ask the philosophical question: does this process eventually converge to a knowledge singularity more affectionately referred to as “absolute truth”? Or is our species guaranteed to tinker at mysteries for eternity, getting ever so close to the mind of God without actually being able to touch it? Did Michaelangelo get it right at the Sistine Chapel?

My own hypothesis is that the pursuit of truth is an asymptotic affair. To substantiate my argument, let me turn to some of the best tinkerers modern science has known: the physicist Murray Gell-mann and the statistician George Box.

George Box has famously said that “all models are wrong but some are useful”. Going further yet, Murray Gell-mann provides an explanation for Box’s observation: “the only viable model of a complex system is the system itself”. Let me leave aside the implications these observations taken together have for Stafford Beer’s “viable system model” and return to the philosophical question of the absolute truth.

Substituting “the universe” for “complex system”, Gell-mann’s observation would seem to indicate that to fully understand the universe, one would have to create the universe, i.e. play God. Lacking this ability, the distillation of knowledge from the cosmological truth will never converge on the absolute truth for a simple reason: even as we move through Roger Martin’s funnel, we have necessarily lost context in distilling a new bit of knowledge. Said differently, the only way to gain specific knowledge is to lose contextual knowledge. If we were to go after contextual knowledge at the largest possible scale just for the sake of snubbing divinity, we would however run into the practical barrier of recreating the very reality we are attempting to study.