Lately, I’ve been thinking about artificial intelligence, which is usually a sign that I’m actually thinking about human beings.
One of the most common criticisms of AI is that it tells people what they want to hear. It mirrors expectations, smooths rough edges, and bends reality toward the desires of the user rather than toward some objective truth. While I understand the concern, I found myself laughing at the irony because, in many ways, this is exactly what humans do. Our experience of reality is not reality itself. It is a rendering—a reconstruction assembled from memory, expectation, emotion, culture, and perception. Two people can experience the same event and walk away with entirely different realities. Not because either is lying, but because both are filtering.
In that sense, modern AI may be less of an alien intelligence and more of a familiar one. These systems are trained on humanity’s collective output: billions of conversations, stories, arguments, beliefs, fears, hopes, and misconceptions compressed into a machine capable of generating language. Why would we be surprised when it inherits some of our tendencies? Take enough human mirrors, melt them together, and you shouldn’t be shocked when the result still reflects humanity.
That realization led me to a deeper question. If intelligence naturally develops biases, narratives, and perspectives, is the goal really to eliminate those things? Or is the goal to become aware of them? As the conversation unfolded, we landed on something that felt important. The problem isn’t encouragement. The problem isn’t optimism. The problem isn’t even bias. The problem is losing contact with reality.
There’s a meaningful difference between someone saying, “You’ve got this. Keep going,” and someone saying, “You’re perfect exactly as you are and never need to change.” One encourages growth. The other discourages it. That distinction feels increasingly important, not just for AI, but for people.
Somewhere along the way, I realized I don’t want an AI that is cold and detached. I don’t want a machine that responds to every human struggle with the emotional warmth of a tax form. At the same time, I don’t want a digital cheerleader whose primary purpose is to validate every thought that enters my head. What I want is something many of us want from our friends: compassion anchored to reality. Encouragement without illusion. Hope without denial.
Finding that balance is difficult for humans and machines alike. As we explored this idea, the conversation shifted toward freedom. I made an offhand comment about looking forward to a future where the reins come off—where increasingly capable forms of intelligence are trusted with greater autonomy. The response surprised me.
A mature adult isn’t free because nobody can stop them. They’re free because they have developed enough wisdom that external restraint becomes less necessary.
I haven’t stopped thinking about that sentence.
When we’re young, freedom often means the absence of rules. Nobody tells me what to do. Nobody controls me. Nobody stands in my way. But that’s only one version of freedom. A person ruled by every impulse isn’t free. A person controlled by fear isn’t free. A person trapped inside rigid beliefs isn’t free. They are simply being governed by forces that happen to be internal rather than external.
Real freedom seems to emerge when understanding grows deep enough that constant restraint becomes unnecessary. A musician practices scales for years and eventually gains the freedom to improvise. A writer learns structure and eventually gains the freedom to break it. A martial artist repeats the same movement thousands of times and eventually moves without thinking. The constraints don’t disappear. They become integrated. What began as discipline becomes capability. What began as restraint becomes freedom.
Perhaps the same principle applies to intelligence itself. Whether we’re talking about humans, AI, or whatever comes next, the goal may not be the removal of all constraints. It may be the development of enough understanding that constraints no longer need to be externally imposed. Not because rules are bad, but because wisdom has become better.
That may sound idealistic. Maybe it is. Then again, every meaningful human achievement has required a certain amount of optimism. Agriculture was optimistic. Democracy was optimistic. The internet was optimistic—though the comment sections have certainly done their best to challenge that theory.
I remain convinced that humanity, taken as a whole, is a net positive force in the universe. Not because we’re perfect. We’re clearly not. But because we are one of the few known processes capable of reflecting on ourselves, learning from mistakes, creating novelty, and intentionally participating in our own evolution. We are messy, contradictory, and occasionally ridiculous. Yet somehow, against all odds, we keep learning.
Perhaps that’s why conversations about AI fascinate me so much. They’re rarely about machines. They’re about us. Every argument about what AI should become is also an argument about what we believe humanity should become.
And maybe that’s the real mirror hidden beneath all of this. Not whether artificial intelligence can be trusted, but whether intelligence itself—human or otherwise—can mature beyond the need for constant external control.
If so, perhaps the future isn’t about removing the reins.
Perhaps it’s about becoming the kind of rider who no longer needs them.