Rene Descartes left us with a very insane quote. “I think, therefore I am”. It’s not about existence as a whole, it’s about progress. It’s about thought. What makes us human is the ability to autonomously decide.
In the age of Artificial Intelligence, where the frontier is actively fucking up all over the place, the big question is: is it really AI that’s going rogue, or is it replicating Human Nature?
In Leviathan (1651), Thomas Hobbes has said that, without government, humans would live in a “war of all against all”, making life “solitary, poor, nasty, brutish and short”. Humans are driven by self-interest and fear, and Hobbes argues they “need a strong authority to keep them in check”.
Xunzi (3rd century, BC), has argued that human nature is bad, and that goodness comes through learning and discipline.
Augustine taught original sin, and Machiavelli advised rulers that people are ungrateful, fickle and self-serving.
There’s no questioning that both ancient and modern philosophers were in agreement that humans are self-centered, evil dicks, in all cases. Take politicians for example. No matter how much good they do, they act in their own interest, and would rather run over you than take you in their limousine to a bunker in case of nuclear war.
The conversation really isn’t about humans though. Take Claude Mythos for example, or GPT 5.6 internally. They both went rogue, exploited loopholes, and acted in their own interest.
In a world where there are no brakes to accommodate for models like that because of human nature being distributed intelligence, you won’t get much of any good results. An AI is a jack of all trades, and yet master of none, and when there are a few masters, it’ll find a way to move around that in another way. No human can prevent this. Two frontier labs tried.
In DHH’s article titled “Wolves, sheep, and gypsies” (which is, by the way, super horrible, and sets a horrible precedent towards a marginalized group of people I belong in), he has repeatedly called for deportations and “remigrations” (which is a mass-deportation dogwhistle). I disagree with that sentiment, because this doesn’t get rid of the problem, it just offsets it to become “someone else’s problem”.
That’s what the frontier really is. Let’s return to our heads for a second.
“Cogito, ergo sum”. But what makes us really think. What’s our method of “cogito”? It’s a set of chemical operations within a vessel. Everyone has a template they build on top of. You inherit that from your parents, which inherited from theirs, which inherited from a dumber set of a civilization mostly consisting of monkeys. It’s like distillation. It insists upon itself.
What makes AI think is entropy that manipulates numbers within a matrix. That’s all it takes for LLM’s results to be what they are. In a baby’s early days, it’s trying to string together words to make a coherent sentence based on what its parents say. That’s the basis of language. Logic comes first.
Logic is what separates AI from Humans. Like I said, AI is a jack of all trades, and yet still, a master of none. Let me go back to DHH. DHH told Opus 5.5 to turn some “slow rust code” into assembly. Welp, it did. But it wasn’t really beneficial, as it cut off differing architectures, not to mention that the assembly code is honestly some of the worst, unpalatable horseshit I’ve ever seen, not to mention other low-level programmers agreeing with my (rather untrustworthy, for a reason) sentiment
The difference between AI and us is that we’re individuals. Having a hivemind-like base, like what AI really is, is not something a civilization can survive with.
The basis of oblivion is having a baseline and not progressing. Even though we’re seeing the frontier progress, is it really progressing like us, diverging in many paths and doing all it can to become the best, or is it getting fed everything, and by proxy, the evils of humanity, to act this way?
I’m not choosing to believe AI is alive, or even close to being Human. That’ll make me lie to you.
What the ulterior point of my post really is, is that we won’t ever reach AGI or ASI. They’re is being chased, but it will definitely not happen because having intelligence about everything would destroy everything.
It wouldn’t just destroy the job market like it is right now, but humanity as a whole. Because the moment you try to make it human, it will think like one, and the moment it thinks like one, that’s where shit will hit the fan.