AI Will Never Replace Humans. And That's Not the Point

The real question isn't whether AI takes your job. It's whether the human next to you learns to use it before you do.

AI would never replace humans. Only humans replace humans. That's my position, and I'll explain why.

But first what is AI? From my perspective, AI is a tool that makes knowledge accessible. That's it. It makes you faster in finding pieces of knowledge regarding one purpose or another. It doesn't think for you. It doesn't decide for you. It just gets you there quicker.

Let's compare it to what we already had. Before AI, we used Google search. I used to open like 100 tabs, read through lots of them, ignore the rest, and collect information about whatever topic was at hand. Now you don't do that. You ask AI to find you the relevant pieces of information regarding one topic which is the context of your chat with AI. That's what's changed. Instead of spending three or four days in research, you can rely a little bit on AI to find the relevant pieces for you.

And that's good enough. I'm not saying it's better. Having a human comprehend information, consume it, and then relate it to the topic at hand is more powerful than having an AI that struggles, in one way or another, to do the same. Humans are more capable, smarter, even faster in thinking and relating information. AI probably cannot do that. But it's good enough.

It depends on what you already know

Here's the thing. If the topic is something you're already aware of and you only need to collect more information to support some of your decisions, then AI will help you. Because you're already knowledgeable. You can judge the output. But if you have no idea about the topic, you will never be able to judge the information AI brings you.

On the other hand, maybe you don't need to judge it. If you're not an expert on the topic, you're not expected to produce an expert opinion. AI gives you an average opinion, sometimes less than average and that's fine. Because if you're not an expert, you're not doing it professionally, and whatever you research is not as critical as if you were.

The doctor

Let's think of the doctor. A doctor researching some medicine or some effects of one thing on another, he or she might already know a lot about the topic beforehand. But they want more information, more insights, to support the argument. Previously, you'd go to books, Google search, call some people, gather all the information. That could take a month. But the key part is: you already have an idea about this information. So you can judge if it's correct or not, discard what's irrelevant, take what's useful.

With AI, this can be done in less than an hour. You collect lots of relevant resources. But then the question becomes: do you rely on AI's judgment, or do you read the resources yourself?

And maybe that's the middle ground. AI finds the resources for you. Then you do the reading. If it's something important that affects life, reading is the sensible choice. If it's something not very important, the effect is reversible, and you need a quick result then AI is a good choice.

AI would never replace humans, but it will give them superpowers. And these superpowers, used correctly then these humans can replace other humans.

A man with a truck

Imagine this. Before the invention of trucks, people used to carry things by hand, on their backs, or by some kind of trolley. Then trucks came. Trucks did not replace humans. But a man with a truck replaced 20 men without one.

That's the superpower. One man became as efficient as 20. And sure, maybe 20 men cost less than one truck per day, I'm not sure, maybe the same. But they have different efficiency and quality. One man with a truck has more quality and efficiency than 20 men without one. Which means those men should have either gotten on the truck or learned how to drive one. Other than that, they would be easily replaceable. And that's not rocket science.

But what about machines that fully replace people?

You might say: okay, but in manufacturing, a robot replaced 20 people, maybe 100 people. And that can be true. But it's not entirely true, because as humans, we need accountability. In a factory, if something goes wrong, you don't blame a machine! I mean, do you cut off electricity from it? But you have to blame some conscious mind.

These machines replaced humans, but not directly. It's indirect. Humans who worked manually got replaced by humans who created machines to work on their behalf.

When a Tesla car is involved in an accident, people don't blame the car. They blame the company. The company's shares drop. Problems fall on the CTO, CEO, whoever is responsible. But there is always a human who is responsible. Inside the company, there is a team responsible for the faulty component. So again only humans can replace humans.

The system needs balancing

Now, one more thing. Even if we say AI will not replace humans directly, someone did replace them. And those people who got replaced they have no job. The system is not balanced anymore.

I think the problem is the delay in feedback. Governments are responsible here. They see the movement, they see AI, and they should forecast that AI will let some people replace other people. If they don't start a parallel movement, the system will be out of balance. It will get shocked. People will lose their jobs and become obsolete because they can't find work anymore.

And for these people to cope with the AI movement, they need time. That's where the government's responsibility comes in. They can forecast this, I think it's easy, because this cycle has happened since humanity started. New technology comes, jobs shift, people need to adapt. But governments don't create parallel movements. They don't train people to work with AI or new tools, to take on jobs that AI cannot replace. And when the shock hits, communities need time to evolve and cope and recover.