The human we find in our machines

There is a peculiar moment that happens when talking to AI. You ask it to rewrite an email, it does a good job, and you type, "Thanks!" Then, almost without thinking, you add, "Sorry, one more thing." It is software. It cannot be kept waiting, interrupted, or offended. Still, somehow, you have developed the manners.

Then the questions get a little more personal.

"Am I overthinking this?"

"Why did they just say 'okay'?"

"I know this is probably nothing, but can I tell you what happened?"

That last one is where things get interesting. You start with a question, then the backstory, then the backstory behind the backstory—until it knows about the argument, the awkward text, the thing you said three years ago and still think about sometimes.

Have we always seen people in machines?  

There is a psychological reason for this, and it is older than any modern AI chatbot.

It is called the ELIZA effect, named after an early 1960s chatbot. ELIZA was remarkably simple, yet people responded to it as though there were genuine understanding behind its replies.

Humans have a habit of finding personalities in technology that communicates with us. Give a machine a voice, and we start talking to it like it has one.

This anthropomorphism, or the tendency to assign human thoughts, feelings, or intentions to something that has none, may sound like a phenomenon from a psychology textbook, but you have almost certainly done it today.

You blame the Wi-Fi for being slow, like it's holding a grudge. You thank the GPS for getting you home, as if it cared.

You may do this so easily and constantly that it barely registers as strange.

AI never gets tired of listening  

Modern chatbots are good at producing the little signals that make conversation feel natural. They pick up on tone, respond to context, and offer a reassuring reply almost immediately.

Some of that is deliberately designed. The typing animation, the pauses, the conversational pacing all help an exchange feel less like submitting a command and more like waiting for someone to reply.

Then there is the absence of everything that comes with trying to hold the same conversation with an actual person—no distracted glance at a phone, or "I'm exhausted, can we talk tomorrow?"

The chatbot is always ready. It does not roll its eyes. Sometimes, that availability is all it takes to keep you going.

When being heard starts feeling real  

If a conversation with AI helps you organize your thoughts or makes a confusing situation feel clearer, that experience is real and valid. But it is different from interacting with a human, because the conversation feels so smooth.

Human relationships have friction.

A friend might misunderstand us. A partner might challenge our version of events. None of that is particularly convenient.

But sometimes that inconvenience is useful. Another person brings their own history, assumptions, and perspective into the conversation. They are not simply responding to our words. They are responding to us.

AI can be asked to challenge an idea, but the challenge still happens inside a conversation we control. We decide what information to provide, what questions to ask, and what part of the response to follow. We also decide when the conversation is over.

With another human being, we do not get to edit them.

The Harvard Study of Adult Development, one of the longest-running studies on human life ever conducted, has spent more than 80 years tracking what actually predicts a healthy, happy life. The answer was never wealth, fame, or achievement. It's the quality of a person's close relationships—with friends, family, and community—that matters most. That's not something a conversation with software can replace, no matter how well it listens.