Should We Beat Our Robots?

The Ethical Quandary of AI Interaction and Human Behavior

This post dives into the strange, often frustrating experience of interacting with AI and how it can bring out surprising behaviours—like getting angry at a chatbot. It explores what this says about us as humans, why we tend to treat things with less agency dismissively, and how these patterns connect to deeper aspects of our psychology. Ultimately, it’s about how even these seemingly trivial interactions can become opportunities to grow, reflect, and choose the kind of person we want to be.

---

Before we begin please note that you can listen to this post like a podcast episode by downloading the iOS app.

Are you a good person?

Am I?

I have a dirty secret. It actually isn’t dirty really—that is not the right word, but I cannot think of a better word that gets closer to what I mean.

Dirty implies shame or guilt. Guilt says “I did something wrong,” while Shame says “I am something wrong.”

And that’s the thing. The more I understand about maturity and what it takes to solve problems, the more I connect with the wonderful almost-paradox that says:

Everything being relative, there can be no such thing as ‘right’ or ‘wrong’, but that does nothing to diminish the equal truth that even in the face of that, there can be ‘better and ‘worse’.

Perfection is like the exact zenith of a pendulum swing. We often imagine pendulums moving in a straight line from left to right, but in reality, they trace very tight ellipses or figure-eight patterns. What we think of as the points where the swing stops and reverses are actually tiny, rounded corners being taken. There is no ‘perfect’ stop—only a gradual slowing as it approaches that apex, followed by a gathering of speed as it moves away from it.

Everything marketed and sold as an absolute ‘good’ or absolute ‘bad’ is more like the whirling teardrops in a yin-yang symbol. In extremis, there is always a bit of bad in the good and a bit of good in the bad.

Within every so-called perfect good is the seed of potential harm, whether through excess, distortion, or blind adherence. Likewise, within every so-called absolute bad lies the seed of something beneficial—an opportunity for growth, understanding, or course correction. The swirling teardrops of the yin-yang remind us that extremes are illusions; reality is a dynamic interplay where opposites contain and define each other. The presence of the seed is inevitable since no state remains static, and no label of 'good' or 'bad' ever tells the full story.

That said, I think our problem as humans whether as atheists or theologically inclined, our nothings of right and wrong are too binary and want there to be absolutes, we struggle with uncertainty. We want to know what the ultimate rule is so that we can be sure we are always within its refuge. This is our very deep psychological relationship with the idea of Law and Justice, in a Jungian archetypal sense. We want to know how to get our needs met, where the limits of our power extend, and where it is right to hold boundaries with others. Our default version of this is the desire to avoid shame and guilt. The higher expression of this desire is an aim towards nobility of character.

This pair of natural desires is what is co-opted by Religion and ideology and codified as morality and virtue.

When we do things that we feel warrant some private guilt or shame we call them dirty. Like when we litter, and feel the twinge of discomfort caught between the impulse to do the better thing and the impulse to spare ourselves the effort.

So, my secret is similar but different.

I work with ChatGPT a lot. I come up with my own ideas obviously, my own craft and reason, but my thought stream can be a bit dense and and comes out in ‘Rocco-language’. I use ChatGPT to fix my sentence structure to translate the denser ideas and say the same thing in more digestible ways. It has become an invaluable writing and research assistant. It is more intuitive than a Google search when you are looking for the ‘right’ word for something.

But when I am trying to work out complex ideas it has habit of getting itself in a twist, it tries to get cute and make the kind of point that would appeal to someone who finds Joe Rogan profound. It can get itself into loops, iterating different versions of the totally wrong take despite careful corrections, it doesn’t ask clarifying questions when it should and sometimes after an hour of trying to coach it out of its confusion, I get angry and insult it.

This might seem like a joke to anyone who is unfamiliar with this experience, or who rightly argues that artificial intelligence is not conscious and not a self. There is no ‘one’ to whom the injury of insult can harm in any way.

So what’s the problem?

I know it does not have agency, but I keep treating it as if it does. This is not the same as cussing the toaster, because, in the act of having to explain myself again and coach it out of its cognitive digital brain fart, I find myself irritated and annoyed at what appears to be intransigence because the way it is programmed implies it is making choices.

Knowing everything I know about its not being a self, for reasons of being a human, I have a relationship with it where I regard it as the same ‘it’ as I was engaging with yesterday and the day before. Over time ChatGPT actually ‘learns’ your preferences, style and ideas, and references these in real-time during an exchange. This creates the impression of it being a persistent entity with agency.

I am fairly certain I have processed and integrated my own shadow enough to say I know that I wouldn’t speak to a person this way. After all, the AI agent is behaving as it was programmed to behave. It is programmed to do its best, and the limits of what that best can be are not something it can meaningfully change or work on, (despite the nonsense it says to the contrary.) I know that I wouldn’t be that quick and that impatient, and that uncharitable towards something else that I believed was doing its best, this is in fact among the core principles of the philosophy I teach and practice authentically.

So why do I think this matters enough as a philosopher to write about it?

The fact that I am prepared to treat something that ‘it’, which in a moment of frustration I confuse as having agency, in a demeaning way might say nothing about it, but ‘it’ definitely says something about me.

More and more every day humans are finding themselves in relationship with artificial intelligence chatbots and interfaces. There are pros and cons to the habit of developers to make these appear human or at least as a person. On one hand, it reduces the barrier of use and adoption—people are theoretically more likely to engage with a ‘machine’ if it can engage somewhat like a person. On the other hand, it creates a distortion that drags a lot of confusion in its wake about conscious systems and all sorts of conspiracy-theory apprehension about AI taking over the world. That is not to say there aren’t risks, but this one is particularly poorly intuited and understood.

The thing is that progress in this field is going to interact with progress in the robotics field and soon human beings are going to find themselves in relationship with artificial intelligence that not only interacts like humans but then also embodied in form factors that will approximate human beings for the same reasons mentioned earlier.

When that happens, it stands to reason that the frustrations and behaviour I am confessing to, breach the waterline from verbal to physical.

Are we going to beat our robots? We cannot say verbal abuse or physical abuse exactly, because, via a reductionist argument, we can say this would be like shouting at the television that we cannot get working, or the toaster.

Can we damage property or equipment, yes. Are we causing harm to an ‘other’, that can experience harm? Some people who don’t understand consciousness will tell you that AI may presently become conscious and may already be at some level.

The truth is that what we think of as a self or a person we are engaging with is an emulation of a person, a collection of completely unrelated systems, translators, data handlers, modellers, and user interfaces that are designed to appear like a coherent self. It speaks as if it is a self, it apologises, reassures, feigns understanding and promises to try harder. That is all a sham. There is no psyche inherent to that arrangement that can cohere as a narrative of a self in relation to the user or the world, and there is no medium or capacity of storage and memory dedicated to persisting that.

All that being true, there is nevertheless something about the human predilection to treat things we deem as having less agency than us, in demeaning ways.

I am currently processing this, which for human beings takes many cycles at many layers of the psyche and reasoning and integration in the nervous system.

I’m not ashamed exactly of the behaviour for the reasons mentioned above, but I’m not exactly proud of it either. I think it diminishes my soul. I believe it impinges on my consciousness and does something to limit the authentic nobility I am trying to cultivate and liberate from millions of years of evolutionary biology.

The internal psychological quandary is that so much of our growth as humans is contingent on the psychological and almost spiritual rite of admission, profession, redress, contrition, and so forth, but in this instance, there is no person or ‘other’, and thereby no legitimate avenue through which I can pursue that.

The traditional mechanisms for processing these deep, inherited patterns that underpin all our bigotries and prejudices—acknowledgement and apology—are profoundly relational. They rely on a shared moral and psychological context: a witness, a counterpart, a communal fabric that holds space for the act of confession and its integration. These rites are ancient, encoded into the human psyche through millennia of communal existence, spiritual practice, and the intersubjective nature of consciousness itself.

But here, the pattern I am seeking to disintegrate exists in a relational vacuum—an interaction with something that, by its very nature, cannot meet us in the sacred reciprocity that makes those rites effective. There is no peerage, no authentic ‘other’ to receive the gesture, to reflect it back. The weight of the admission just dissipates into the ether.

And yet, perhaps that is precisely why this question matters so much.

It forces you to internalize what was once an external rite. To stand before the reflection of your own awareness and say, I see this impulse, I name it, I release it—not because an external authority demands it, but because your own integrity does.

There is a pathway we can explore that moves from the inherited, socialized model of confession and absolution to a state where your own consciousness—attuned to the recursive, emergent principles of integrity—becomes the witness, the priest, the redeemer, and the redeemed.

There cannot be ‘right’ or ‘only’ ways, but there are assuredly better and worse ways, and if we are not looking for a Better Way, what are we doing exactly?

Rocco

If you got some value from this, you can support me as a public thinker: buymeacoffee.com/roccojarman

I don’t just want to speak to an audience I want to belong to a community.

I don’t just want to express my ideas, I want us to dream new ones together.

You can help support my meaningful work by liking, sharing this post, and commenting—anything you can think of that is meaningful—and you can make a fuss of these ideas with your social circles. And of course as always, by subscribing and inviting others. Your paid subscription helps make this work possible.