What makes humans special?

A common argument against AI is that it “doesn’t really think.” I find this strange, because the examples used to prove it are often tasks that dogs would fail too. Does that mean dogs do not think? Obviously not.

Other examples relate to creative tasks, such as “AIs cannot draw because they sometimes depict people with six fingers.” Usually, this is said about an image that looks better than what 99% of humans could paint. Yes, most people know that a hand has five fingers. But most people could not draw a photorealistic hand; instead, they would doodle. So, we ask AI to perform a task most humans cannot perform, then declare it unintelligent because it fails at one small part of the task that most humans would get right.

Or take the other classic example: “AI cannot count how many Rs are in the word strawberry” (they say it’s two, or at least, that’s what they used to say a year ago). If you had asked the same question to an average person 500 years ago, would they have answered correctly? Probably not, as most people were illiterate. That does not mean humans became intelligent only once mass literacy arrived.

This is the problem with many AI intelligence tests: if a test would fail most humans today, or would have failed most humans in history, it cannot be the test that separates humans from machines.

Once we accept that constraint, it becomes hard to deny that AIs are intelligent in many meaningful ways.

So, it’s probably not intelligence that makes us special. Could it be emotions?

I don’t think so. Ask a recent AI model what someone would feel in a given situation, and it will usually answer as well as a high-EQ person would, and probably better than a low-EQ one.

What about creativity?

Again, not really. Some humans are highly creative. Most are not. There is a famous exchange in the movie “I, Robot,” where Will Smith asks, “Can a robot write a symphony? Can a robot turn a canvas into a beautiful masterpiece?” and the robot replies, “Can you?”

So what makes humans special?

I think the answer is not intelligence, emotion, or creativity. It’s scarcity and skin in the game.

When I speak with Claude or ChatGPT, I am not really receiving something scarce. The model can answer me, and you, and millions of others at the same time. But when I speak with my spouse or with a close friend, their attention is scarce. If they are with me, they cannot be with someone else. That is part of what makes it meaningful.

We, humans, are special because we are limited. We have one body, one life, one stream of attention. We cannot be everywhere. We cannot speak to everyone at once. We cannot give ourselves infinitely.

And we also have skin in the game. If a human gives me bad advice, there are consequences. I can stop trusting them. I can damage their reputation. I can sue them, leave them, fire them, or punch them in the face. AI does not face consequences in the same way. It can be wrong, but it does not suffer much from being wrong. It has no body to protect, no reputation of its own in the human sense, no single life that can be damaged by bad judgment.

That, I think, is the difference.

We, humans, are not special because we are the only intelligent beings. We are special because our intelligence is trapped inside a scarce body, moving through a single life, and making choices that can cost us something. That is what makes us special.