For decades, machines operated only by responding to direct user commands. In other words, computers were designed to perform set tasks in response to pre-programmed commands. Now, computers don’t strictly need to receive an ‘input command’ to perform a task but rather ‘input data’. Specifically, the machine creates a predictive model based on previous experiences captured in the data. From the input of data, the machine is then able to formulate decisions on how, where and when to perform a particular action. In the case of our earlier example, the model was predicting whether the pizza delivery would be late based on existing data.

几十年来，机器只通过响应直接用户命令来操作。换句话说，计算机是为了响应预先编程的命令执行设定的任务。现在，计算机不需要严格地接收“输入命令”来执行任务，而是“输入数据”。具体地说，机器基于数据中捕获的先前经验创建预测模型。从数据输入，机器就能够制定如何、何地和何时执行特定操作的决策。在我们前面的例子中，模型是根据现有数据预测比萨饼的交付是否会迟到。