I read Meditations by Aurelius about a decade ago, before the bastard Ryan Holiday ruined all discussion around Stoicism. Firsts are always special, and Meditations was the first philosophy book I picked up on my own. I revisit it every few years, and each rereading leaves me with a different line that sends me on a rabbit chase. Last time it was: "If it is not right, do not do it; if it is not true, do not say it." The "do not lie" part was easy enough to understand, but what exactly counts as "right" was much less obvious. Looking for an answer, I quickly stumbled upon the categorical imperative, which seemed to be exactly what I was looking for.

The gist of the book is as follows:

Kant begins by distinguishing between rational moral philosophy and empirical moral philosophy, the latter being based on a posteriori principles and the former on a priori ones. His focus is on rational moral philosophy--how people ought to act independently of experience.

Kant proceeds by arguing that the good will is the only thing that is good unconditionally and that only actions performed from duty possess genuine moral worth. Moral worth comes from the maxim behind an action, not from the purpose or consequences the agent hopes to achieve. Simply conforming to the moral law is not enough; one must act out of respect for it. This leads to the first formulation of the categorical imperative: "I am never to act otherwise than so that I could at the same time will that my maxim should become a universal law."

Kant develops the categorical imperative through several formulations:

The Formula of Universal Law
"Act only on that maxim whereby you can at the same time will that it become a universal law."
The Formula of the Law of Nature
"Act as if the maxim of your action were to become by your will a universal law of nature."
The Formula of Humanity
"So act as to treat humanity, whether in your own person or in that of any other, in every case at the same time as an end, never merely as a means."
The Formula of Autonomy
"The idea of the will of every rational being as a will giving universal law."
The Kingdom of Ends
"Act according to the maxims of a member of a merely possible kingdom of ends legislating in it universally."

These formulations have overlap and together they provide a robust framework for evaluating not merely actions but the motivations behind them.

Kant concludes by arguing that because human beings are free and rational, they are morally self-legislating. Our dignity lies in our capacity for autonomous rational agency, and acting in accordance with moral law is an expression of that dignity. Morality is therefore not about maximizing happiness or obeying an external authority, but about freely governing oneself according to principles that every rational being could endorse.

I found the Formula of Humanity to be the most compelling formulation. I see the Formula of Universal Law as a test for uncovering what genuine respect for rational agency requires. Take Kant's example of borrowing money with no intention of repaying it. If false promises were universalized, the very institution of promising would collapse. The universalization reveals the violation rather than creating it. The deception manipulates another person's rational agency by preventing them from making an informed decision, effectively recruiting them (means) as a tool in your own projects (ends) without giving them the opportunity to opt out.

Recently, the algorithm has been feeding me videos of people letting AI guide every aspect of their lives--asking it what to wear, what to eat, who to date, what career decisions to make, and what opinions to hold. It made me think about Kant's conception of agency. Extending Kant's framework to the use of AI, the gradual delegation of judgment erodes the exercise of moral agency. For an action to possess moral worth, it is not enough that it merely conforms to the moral law; it must also be performed from duty by a rational will. Decisions delegated entirely to AI therefore lack any moral worth. Taken further, habitual reliance on AI risks replacing autonomous judgment with convenience. Sure, AI could be a capable substitute for practical reasoning, but in doing so it would gradually diminish the very capacity that, grounds our dignity.

Also, I recently read The Gay Science, so Nietzsche calling Kant "Chinese" was still fresh in my mind. First of all, to my cumtown–ruined brain, it's a funny image. But, I understand what Nietzsche was getting at. Compared to Nietzsche's poetic prose, Kant's relentlessly systematic style feels lifeless. But in Kant's defense, someone had to attempt a rigorous account of pure moral philosophy, and the precision of the project justifies the dryness of the presentation.