Why the Mirror Test Fails: What a Cat Reveals About Consciousness and AI
Hey, I been somewhat slacking with my blog, as I have ‘launched’ re-run, the sci-Fi novel im writing. I publish the chapters in a separate blog, so this doesn’t become to messy. I noticed, I should post stuff like this in a separate format, maybe. Anyhow, next one will be.
All the hype around Artificial General Intelligence (AGI) won't get us any closer to one thing: a true understanding of consciousness. And that's the crucial, missing piece that we simultaneously know everything and nothing about.
But what is consciousness, really? Is it just the realization of self?
I think (I comprehend my existence), therefore I am (conscious)?
For decades, science has relied on a seemingly simple tool to answer this: the mirror test. The concept is straightforward: place a mark on an animal's body and see if it recognizes the reflection as its own by touching the mark on itself. If it does, we tick the 'self-aware' box. But is it really that simple?
The Limits of a Reflection
The problem with the mirror test is that it contributes a single action, touching a spot, to the vast, complex concept of self-awareness. It assumes a conscious, deliberate choice. But what if the action isn't a choice at all?
What if it's just a sophisticated reflex? This is where we need a different perspective.. While there's likely a scientific term for it, perhaps something related to empathy, it needs a name for our purposes. So, for the sake of this argument, let's call it the 'Generalized Extended Cat-Button Theory'. I feel the word 'Extrapolation' is missing, but I'll spare you for now.
Cat-Button Theory
To get behind the concept of GECBT you first have to understand the (simple) Cat-(lick)Button Theory. In simple terms, the theory predicts that every type of cat has (lick)Buttons placed at random points on their spine, up to the beginning of the tail.
It also projects, that if there is a cat, with no apparent (lick)Button, it has it’s first theoretical occurring (lick)Button behind it’s actual size (it’s to small to have it). When these nerve-dense regions are stimulated, they trigger a specific, involuntary response, often a lick. Whether you see this as a direct reflex or a form of "displaced behavior," the critical point is that the action is widely considered involuntary.
So, when an animal in the mirror test reaches for the painted dot, are we witnessing a profound moment of self-realization? Or did we just unknowingly press a neurological 'button' that triggers a seemingly intentional action?
The Brain as a Storyteller: Our Own Justification Module
Before we dismiss this, consider our own brains. We've all experienced something similar. Think of that moment when you're drifting off to sleep and your body suddenly jolts awake. If you fully wake up, your brain, a master storyteller, has often already invented a reason. I, for instance, have woken up from this convinced I was dreaming of running on a railroad and the kick was me tripping over a railroad tie. This is our 'justification module' at work, creating a narrative for a physical event it doesn't initially understand. It proves that even for humans, the line between an action and a conscious reason for it is blurry.
Sunfish, while typically preferring to live solitary lives, are naturally curious and tend to approach divers in the wild.
X / @shimonoseki_aq
This relentless focus on self-recognition also misses a more fundamental point, a point perfectly illustrated by a lonely sunfish in a Japanese aquarium. When the aquarium closed for renovations in December 2024, the sunfish became so depressed from the lack of visitors that it stopped eating. The staff's ingenious solution? They placed cardboard cutouts of visitors in front of the tank to cheer it up.
This raises a crucial question: does it matter if the sunfish can recognize its own reflection? It can clearly feel sadness and, by extension, probably depression. Isn't the capacity for suffering and joy a far more profound indicator of a rich inner life than simply passing a visual test? Maybe consciousness isn't the right metric; maybe it's the subconscious that's truly in control.
I guess if that Sunfish would be a ‘special’ Sunfish, it would not only still be depressed, but it would also develop some severe paranoia.sunfish. X / @shimonoseki_aq
Why True AGI Is Still a Pipe Dream
And this is why the path to AGI is far longer and more complex than its proponents admit. We are pouring billions into creating artificial minds, yet we're still using rudimentary tools like the mirror test to understand the natural ones.
If we can't definitively distinguish a moment of profound self-awareness from an involuntary twitch in an animal, and if our own brains invent stories to explain our reflexes, how can we possibly hope to build or even recognize true consciousness in a machine? By some definitions, we are close to AGI, and that may be true. But if you call that AGI, I call my blog the successor to Schopenhauer’s “The World as Will and Representation”.
For me, this isn't just a philosophical exercise. Since the ‘fog’ in my brain cleared, I've gained a much better understanding of these mechanisms in my own life. It feels as if my subconscious was protecting me from something. That’s also why I'm rather obstinate with certain ideas; I've been following the same suggestions forever, and they didn't work. Now, I have opened my eyes and I see. Whether I like it or not.
