The Ukraine of Why
Do no harm
Bootstrapping a compiler is a non-trivial test of the language being compiled, and as such is a form of dogfooding — eating your own dog food — the practice of consuming one’s own wares.
Take, for example, the Kyiv manicurist who during blackouts paints her own nails at state-sponsored “invincibility points" with electricity produced by donated generators. Power outages did not cause her self manicures, even though they have become highly, highly correlated since early in October.
Other examples of dogfooding in big cities are abundant. One of my favorites in Kyiv takes place in the morning of the last Tuesday of each month in the center of town. That’s when a team of municipal workers rides around in a truck with power tools screwing new bicycle-lane barriers into the asphalt. Very few people ride bicycles or scooters in Kyiv during the winter and the gut cause of extensive bike-barrier damage is unknown to me. But I suspect many are deliberately broken in order for them to be replaced. A dogfooding loop, as it were.
Which brings us to the democratization of causality.
Physics, which is based on Algebra, doesn’t distinguish between these two equations, but the Causal Model does, because it differentiates between different types of pertubation and considers not only possibilities contained in data, but those triggered by data.
We can extrapolate the model to explain the not-so-surprising resignation of Ukraine’s Deputy Defense Minister Vyachoslav Shapovalov his week after media reports caused a stir, alleging his department signed contracts with a company that was spending way too much money on eggs, deliberately, ostensibly to line someone’s pockets. Not exactly dogfeeding, but almost.
Reasoning with uncertainity (war) isn’t the same as reasoning with cause and effect, which is a far more complicated task. Reasoning with intervention (“What if we stop installing new bicycle-path barriers every month?”) and introspection (“What if we had not given all the military’s egg procurement contracts to a murky middleman?”) must invoke causal models, like “Where are we going?” and “What is the next step?”