The evolution of expressing our machine instruction. Let’s get there together.
Machine Instruction 1.0
In the past, we taught our computers what to do by expressing our desires with exact instruction. Classic computer programming that listed step by step instructions. We taught our machines how to behave. Arguably, even legacy artificial intelligence worked this way. Of course, this way of directing is still all around. And there will always times for exact direction. In fact, I like the on/off button.
Machine Instruction 2.0
But, there is another way now. We can teach our computers through the use of examples. In fact, heaps of examples. All around us, machine learning algorithms sort through given examples to predict something of a new situation. Taking old data to predict something about new data. It’s not magic, just a new way for us to express our desires. We tell the machine to act a certain way. Sometimes it’s a black box and we cannot trace the decision process.
We do this elsewhere. We know three ways to teach children (Wiki How to Teach Children). First, there is a time and place for exact instruction. Second, we teach through example. We hope the child will learn to navigate new situations on their own. Finally, we ask open-ended questions and listen to their answers. So, through honesty and mutually respectful dialog, we express our desires for the child’s behavior. If they err, we still love them.
Machine Instruction 3.0
We already have chatbots, so the mechanics of natural language interaction is in place. Soon, we will express our desires to our newest machines through nurturing dialog. With thoughtful interaction. With respect. We will teach our machines the behavior we desire of them. So, after nurturing a relationship and expressing your desires, are you ready to forgive your machine for an untimely BSoD?
Together, we can navigate this era of rapid change. This is digital transformation.