What Is AI?

Exploring these questions will help us think rightly about AI. So let’s find out some answers.

Introducing AI

AI systems are computer programs that seem to act or “think” like humans in some ways. These programs work a bit like our brains. God designed our brains to think using billions of tiny cells known as neurons. These neurons receive, process, and pass along data from our senses.

(Image by NIH Image Gallery via Flickr)

For example, our senses might send our neurons data about the sight, sound, and smell of a crackling campfire. Our brains can then tell our bodies how to respond. Accidentally touching a fire, for instance, will trigger your brain to pull your hand back.

Similarly, AI systems receive, process, and respond to information. But instead of using brain cells,1 AI uses “artificial neurons.” Unlike physical cells, artificial neurons are sets of rules and data for doing specific calculations. These calculations let the system process information. Along the way, the system “learns” by using algorithms.

What Are Algorithms?

An algorithm is a way of reaching a certain goal by taking certain steps in response to certain circumstances. For example, say your goal is to bake a cake. Imagine finding instructions that tell you different steps for how to bake the cake, depending on the circumstances of what ingredients you’re given.

The instructions might say, “If you’re given a lemon, use two teaspoons of its juice. But if you’re given cocoa instead, use a half cup of it. And if you do use cocoa, subtract a half cup of flour from the recipe.” These instructions give you an algorithm, telling you what steps to take under what circumstances to reach your goal.

(Image by Vika Fleysher via Unsplash)

Now, suppose that every time you bake, you learn ways to make your next cakes even better. If your first cake burned, you might decide to remove your second cake from the oven sooner. You’d change your future actions based on what you learned from past experiences.

Similarly, an AI system can learn to change its future actions based on past experiences. But instead of baking cakes, the system’s goal is usually to sort information or make guesses. To see how these guesses work, let’s investigate a common type of AI: chatbots.

How Chatbots Guess

Chatbots are AI systems that seem to answer questions like humans do. To give their answers, chatbots guess the best order of words to put together in a sentence.

(Image by Solen Feyissa via Unsplash)

How do chatbots make these guesses? First, the system studies tons of “training data,” such as words that humans wrote. As a result, the system learns patterns of relationships between words. This learning lets the system guess how to write new sentences.

Making these guesses takes many steps. Not even scientists who program AI can keep track of all the steps a chatbot takes to answer a single question. So we don’t fully understand how chatbots decide what to answer. This makes it hard to double-check whether chatbots are making the best decisions.

How Is AI Different from Humans?

Chatbots often sound human because they’ve learned from patterns of words that reflect the emotion, thought, and humor of human writers. But chatbots can’t feel or think like us. They’re just computer programs.

On the other hand, humans are living beings who God created in his image (Genesis 1:27). God gave us minds that can know him, spirits that can worship him, and hearts that can love him. He gave us bodies that can do his work while enjoying his creation. And he gave us souls that last forever.

(Image by Katherine Hanlon via Unsplash)

God also gave us the ability to be aware of ourselves. We realize that we’re living, thinking, feeling beings.

But is AI alive or aware? Nope. AI doesn’t know it’s processing data anymore than rocks know they’re hard.

AI can string together words that sound like human prayers, poems, or love letters. But these words don’t come from human souls that can form relationships with God and others. Unlike humans, AI can’t really pray, worship, or love.

Truths to Remember About AI

Knowing these facts helps us understand what AI can and can’t do. AI can function as a helpful tool for many tasks. But AI can’t fulfill roles that depend on being able to love others. For example, a chatbot can’t truly serve as a parent, pastor, or friend.

AI also can’t solve the world’s biggest problem: sin. In some ways, AI may help make life in our sin-broken world easier. But AI can’t redeem us. Only Jesus can.

Another important truth is that AI isn’t a reliable information source. AI learns by studying human writings that may have mistakes or lies. Sometimes AI makes up false information. And people can use AI to make pictures or videos that look real but are totally fake.

For all these reasons, we should never automatically believe AI. Instead, our highest authority for truth is the Word of God, who never lies (Numbers 23:19; Titus 1:2).

(Image by Samantha Sophia via Unsplash)

Summing Up

Ultimately, AI systems are computer programs that can do countless things, from writing sentences to creating videos. These programs work by studying data, figuring out patterns, and using that knowledge to perform tasks (like guessing how to put together words). Yet because these programs aren’t alive or self-aware, they don’t realize what they’re doing.

AI can sound human. But unlike us, AI can’t feel, love, relate, or worship. Unlike the Bible, AI doesn’t always give true answers. And unlike God, AI can’t solve our greatest problem. Knowing these truths helps us think rightly about AI in a changing world.

Footnotes

1. At least, most of today’s AI systems don’t have brain cells. But some scientists today are building computers that do process information using real brain cells grown in labs.