FUTURE PREDICTION SYSTEM is a reflection on belief systems that, despite lacking logical causality, are often treated with blind faith based on inductive reasoning. Through scientific observation, we have come to understand and live by universal laws—such as gravity—that appear consistently true over long periods of time. These laws are founded on repeated, exceptionless observations. However, the process of connecting observed data with outcomes sometimes lacks a rational or causal link.

For example, witnessing a meteor fall is an observable fact, but assuming that someone close to me will be harmed as a result has no logical foundation. Nonetheless, historical practices such as astrology and tarot have functioned effectively within societies, even without clear causal explanations. These systems rely heavily on inductive data built up over thousands of years, yet cannot explain why certain outcomes follow specific signs.

In contrast, I have approached the idea from the opposite direction. I artificially generate factual, inductive data and intentionally link it to the future outcomes I want to predict. By observing and recording the state of a large flag swaying outside my window, I collect data, then cast various forms using baking ingredients at that specific moment. As I accumulate this data, I hope for an inexplicable correlation to emerge between the state of the flag and the forms created—an irrational, yet consistent system of cause and effect.

Using this system, as long as I can observe the flag’s movement, I can theoretically predict what form will be cast inside the machine. If the predictions don’t align with the outcomes, it may simply be because I haven’t yet gathered enough inductive data. Consider the Eastern philosophy of saju (Korean Four Pillars), a system based on thousands of years of accumulated data that appears to offer accurate life predictions. Compared to that, I’ve only conducted a few dozen observations—clearly insufficient for statistical reliability.

Yet theoretically, the system isn’t flawed. And I don’t bear the responsibility of proving its causality. I only need to believe in its operation and continue building the dataset. In doing so, I move closer to constructing a FUTURE PREDICTION SYSTEM with a 100% success rate in forecasting the future.