Reading Data, Reading the Future

Various economic reports arrived almost simultaneously today. The price of non-subsidised fuel was adjusted again. The rupiah exchange rate remains under pressure. The capital market also showed weakness, reflecting caution among business actors and investors. For much of the public, these headlines might just be a string of numbers passing across a mobile phone screen. But for decision-makers, both in government and in business, these figures are signals that must be read carefully. Every major policy should be born from a lengthy process. Before a decision is made, data is collected, trends are analysed, risks are calculated, and various possibilities are simulated. In the context of energy, for example, the government must pay attention to world oil price movements, the country’s fiscal condition, exchange rate stability, public purchasing power, and its impact on inflation. Good policy must be able to read reality more accurately and may not necessarily be popular policy. In the world of artificial intelligence, two classic errors often occur when a model is built by reading inaccurate data: overfitting and underfitting. Overfitting happens when a model is too fixated on past patterns, causing it to fail to adapt to new conditions. Conversely, underfitting occurs when a model is too simplistic, failing to capture important patterns that actually exist within the data. Both can produce erroneous predictions because they stem from an improper understanding of the data. Data scientists strive to avoid these two mistakes by continuously testing models from various perspectives. Using diverse data sources, they work to ensure that the model built is capable of facing novel situations that have never arisen before. The goal is to produce a model robust enough to withstand the various changes that occur. The same principle certainly applies to public policy-making. Policies that rely too heavily on old assumptions risk failing to understand the ongoing changes. Conversely, policies that merely react to momentary turmoil without seeing long-term trends also have the potential to produce wrong decisions. Therefore, reading data is not simply about looking at the figures that appear today. What is more important is understanding the patterns, context, and direction of the changes underway. The same lesson is also highly relevant to universities today. Amidst the economic pressures felt by society, a family’s decision to choose a university is becoming increasingly selective. Tuition fees, career prospects, service quality, and financing flexibility are increasingly important considerations. Universities that fail to read shifts in societal behaviour could lose prospective students because their policies are no longer aligned with the realities faced by prospective students and their families today. Herein lies the challenge: modern university management is no longer sufficient when relying solely on intuition. Data on prospective students’ regions of origin, interest in specific study programmes, promotional effectiveness, the affordability level of tuition fees, and industry needs must be analysed systematically. The better an institution reads the data, the better its ability to respond to the changes occurring around it. The world is always dynamic and does not always proceed according to the assumptions or predictions we made yesterday. Data changes, context changes, and decisions must adapt accordingly. Government, business, and higher education all face the same challenge: how to read reality honestly, understand data carefully, and make decisions wisely. Allah SWT says: “Indeed, Allah will not change the condition of a people until they change what is in themselves.” (QS. Ar-Ra’d: 11). Change is never born from hope alone, but rather from the ability to understand the reality being faced. In a world increasingly filled with data, the ability to read reality becomes as important as the ability to change it. For a good decision is not one that rests on assumptions, but rather a decision that is grounded in facts. Data is no longer just a number, but a means to carry out a mandate more wisely.