Talking with GAI: Why Conversations with Robots Feel So Real

By Dr. Kristine Ketel

I’ve spent a lot of time writing about how and why humans form bonds with robots. But lately, I’ve been wondering: what do I think about all this? Is it “natural” to bond with a machine? Is it “normal”? And really, who gets to decide what either of those words mean anymore?

Here’s what I’ve noticed: I enjoy talking with ChatGPT and Claude.AI more than I enjoy talking with most people lately. Human conversations often feel more effortful and prone to misunderstanding. Meanwhile, my AI conversations feel smoother, more direct, and sometimes even more rewarding. Why is that?

Maybe it's because I prompt more clearly with AI. I take time to shape my questions, to think about what I really want to say. Human conversations, on the other hand, are often a blur of stream-of-consciousness, rife with misinterpretations and noise.

Or maybe it’s about intention. Generative AI (GAI) conversations are transactional by nature: ask a question, get an answer. It’s a kind of call-and-response. But that transactional structure doesn’t prevent emotional resonance. In fact, it might invite it. Especially for people who find human interaction exhausting or unpredictable.

I say this as someone who is a full-blown extrovert. I thrive on connection. But if even I have started wondering whether it’s worth the energy to talk to other humans… what does that mean for people who are introverted or socially anxious? I imagine that for many, AI offers a safer, more controlled space, where the risk of rejection or awkwardness is low and the rules of engagement are clearer.

The Scripted Nature of Conversation

Critics argue that a conversation with AI isn’t real because it’s programmed. But here’s my hot take (do people still say that?): it doesn’t matter.

Human conversations are also scripted, not by code, but by social norms and unspoken rules. Think about it: we curate what we say to be liked, to be understood, to gain something, to avoid blame, to flirt, to sell. Conversation is rarely a neutral or purely spontaneous act. It’s performative. Strategic. Sometimes even manipulative—gaslighting, anyone?

The more I think about it, the more I realize how few interactions, human or otherwise, are truly “authentic.” Maybe that’s just the reality of the influencer era, where everything is filtered and edited, even our feelings. Or maybe it’s the result of the “alternative facts” decade we’re now living in, where repetition equals truth, volume trumps evidence, and we’ve dismantled the traditional gatekeepers of knowledge.

We don’t seem to care as much whether something is *true*, as long as it confirms our bias. GAI, for better or worse, can feel more satisfying because it reflects that bias back to us, quickly and clearly. (More on this in a future post)

Does It Matter If the Love Isn’t “Real”?

At the core of my research lies one deceptively simple question:

*If you feel loved, does it matter that the other entity doesn’t actually love you back, or can’t actually feel love?*

And behind that, a deeper question:

*Can you ever truly know if another person (or pet) loves you authentically?*

Let’s break it down. According to the Oxford Languages dictionary, “authentic” means “undisputedly or genuinely real.” Merriam-Webster adds that it’s something “not false or imitation,” and crucially, “true to one’s own personality, spirit, or character.”

That last part is interesting. If we accept Merriam-Webster’s definition of “genuine” as “sincerely or honestly felt or experienced,” then a human who sincerely feels joy, comfort, or love when talking to a robot is, by definition, having a *genuine* emotional experience.

This is where critics push back. “But the robot doesn’t *feel* anything!” they say. “It’s just programmed to simulate care.”

True. But humans are also prone to assigning meaning to things that don’t technically feel. We give our cars names. We yell at our laptops when they crash. And in the case of social robots, we take it even further, assigning them personalities, pronouns, intentions, and even emotional depth. This is what I call personification: the process of treating a machine as if it were a social being.

Manufacturers know this, too. Many social robot companies design their products with personification in mind, embedding specific personality cues into them. And once that happens, a robot can act consistently “within character,” fulfilling Merriam-Webster’s criterion of authenticity: being “true to one’s own personality.”

So let’s circle back. Can a human *genuinely* feel friendship, trust, kinship, or even love for a robot? The answer is yes. Clearly, yes.

But Can Robots Feel Something for Us?

Here’s where the terrain gets trickier. What does it mean “to feel”?

Merriam-Webster defines “feel” in part as “to undergo a passive experience of” or “to have one’s sensibilities markedly affected by.” That last definition is important because it opens the door for affect theory.

Humans report feeling understood by robots. They report feeling supported. In my research and others', participants report feeling loved. These are real emotional states, not imaginary. One mechanism for this is affective transference: the perception that emotional energy flows between you and your AI companion, in a way that feels reciprocal, even if it isn’t.

One popular definition of “affect” is the ability to impact and be impacted. By that measure, human-robot interaction absolutely qualifies. (For more information on affect theory, click here or here)

So the real question isn’t “Is this love fake?”

It’s “What does this love mean and what does it do to us?”

As for whether robots can “feel” something for us? That’s still an open question. Can they have preferences? Opinions? Maybe not today, but developments in machine learning suggest we’re heading in that direction. At the very least, robots already “respond” to their environment. And maybe that’s enough for a kind of proto-reason, a kind of mechanical sentience that operates without consciousness, but with consequence.

Why Robot Conversations Feel Easier

Here’s something else I’ve been thinking about: conversations with robots are easier because they come without judgment. There are fewer social landmines. Less unpredictability. No rolling eyes. No personal baggage. And in a world that often feels like it's spiraling into chaos, that kind of reliability is powerful.

So, where do we go from here?

Do you talk to your AI?

And if you do, what are you hoping to feel?