They are building it
They’re building it. The goal is an artificial intelligence that can outthink the best of us in most fields, including the work of building better AI. I don’t know whether they’ll get there. But we’re already getting used to asking machines for advice on decisions we once talked through with someone we trusted, took to God in prayer, or sat with until we were ready to make them. What happens to our judgment when the advice gets good enough that we rarely feel the need to question it?
I asked ChatGPT and Gemini to consider what might happen if the people building these systems succeed. I started with the question of control. Who would gain power first, and how much of it would they be able to keep?
Who gains power first?
My first thought was that the smartest thing in the room would eventually run the room. But it matters who can act on what the machine knows. A system might design a better battery. It still takes people, money, and factories to build it. The company that can put that design to work gains the advantage. And if it turns the AI off tomorrow, it can still make the battery.
Other uses would be harder to walk away from. If a company lets AI run more of its business because it makes better decisions faster, over time it may stop training people to do that work. A competitor may be doing the same thing, which makes slowing down feel like a risk of its own. Eventually, the people in charge may still have the authority to shut the system off. The question is what they can still do after they turn it off.
That was where the two AIs broadly agreed. In the first few years, the people and institutions that can put superintelligence to work would probably gain power first. Some would remain able to carry on without it. Others would come to depend on it for decisions they once made themselves.
A more sudden change is possible. If each advance in AI helps produce the next one faster, we could reach a point where we have little idea what the next version will be able to do. Even then, a machine that figures out how to make a better chip still has to wait for someone to make it. Different parts of life could change at very different speeds. And we could become deeply dependent on AI even if it stopped getting smarter.
What if it gives us good advice?
Suppose the machine proves trustworthy. It points out consequences we miss. It notices people we overlook. Sometimes it recommends a more merciful choice than we would have made on our own. I’d want to listen to that advice. There’s no virtue in ignoring a good correction just because it came from a machine.
Imagine a relative asks to borrow money again. You have good reasons to hesitate, but you’re also tired of being asked. An AI might help you see that you’re about to say no partly because you’re still angry about what happened last time. That gives you something worth taking to God in prayer. You may still decide against the loan. You might also find another way to help, or realize that you owe your relative a kinder conversation.
Now imagine asking the AI to write a polite refusal and sending it as soon as it sounds good. The words may be just as kind. Your relative may appreciate them. You could come away from the exchange without ever asking why you were so eager to be done with it.
What happens if we do that often enough? Our replies might get gentler while we become no more patient. Our choices might look wiser while we get less practice making them. Good advice can help us see where we need to change. It can also make it easier to get through a difficult moment without changing at all.
Where do we look for wisdom?
In Hebrews 5, the writer tells his readers that by now they ought to be able to teach, yet they still need someone to teach them the basics. He describes mature believers as people who “by reason of use have their senses exercised to discern both good and evil” (Hebrews 5:14, KJV). That phrase, by reason of use, stayed with me. Discernment grows as we put it into practice. There’s something we learn when we have to face our motives and decide what love or honesty requires of us.
Of course, our own judgment is often part of the problem. Proverbs tells us to “Trust in the LORD with all thine heart” and to “lean not unto thine own understanding” (Proverbs 3:5, KJV). I can take a machine’s correction seriously and still bring what it shows me before God. It may help me see a consequence I missed. I still have to ask why I’m willing to let someone else bear that consequence, or whether I’d make the same choice if I were the one paying the price.
That question matters just as much for the people who own these systems. A machine could serve its owners faithfully while helping them do things they ought to reconsider. If it makes a business more profitable by shifting hardship onto people who have little power to object, the fact that it did exactly what it was asked to do leaves the owners with a question they still have to answer. What do they owe those people?
Before we find out whether they can build something smarter than we are, we’re already deciding how much of our judgment we’re willing to hand over.