Some Links, 9/20/2025

Josh Zlatkus on rationality and having children; Ethan Mollick on using AI; Aaron Renn on Gen Z preferences; Holly Mathnerd on Cluster B

Josh Zlatkus writes,

Humans use at least two forms of logic, or rationality, when making decisions:

• Academic rationality is the cognitive technology developed by thinkers throughout the ages, represented by frameworks such as the scientific method, formal logic, probabilistic or Bayesian thinking, Homo economicus, and so on. This kind of reasoning is the Enlightenment’s pride and joy, and typically what people mean when they say something like “be rational.”

• Ecological rationality refers to “fast and frugal” heuristics that humans evolved over their long evolutionary history. By and large, these are well-suited to the tasks and environments that regularly occurred in ancestral conditions, such as: choosing a mate, deciding what to eat, selecting friends, reciprocating favors, getting back to base, and so on.

Academic rationality sounds like Kahneman’s System 2, and ecological rationality sounds like his System 1.

The post is mostly about the decision to have children, which he argues is not handled by either system.

From an academic perspective, incentives drive behavior, so lowering the cost of parenthood should increase births. But it doesn’t, because people generally don’t factor these kinds of incentives into their decision. They rely on a different rubric, a different rationality.

For most of human history, that rubric was simple: Do I want to bang this person? If so, and a child resulted, the next question was: Do I want to protect this offspring? At no point did anyone have to ask whether to have children. At no point did they have to list their values and imagine “living into” them. Reproduction just happened, without deliberation.

Here is how I think about the decision to have children. Everyone I know in my age group would rather be a grandparent than not. So think about how your decision about having children affects your chances of becoming a grandparent.

Ethan Mollick writes,

for an increasing range of complex tasks, you get an amazing and sophisticated output in response to a vague request, but you have no part in the process. You don’t know how the AI made the choices it made, nor can you confirm that everything is completely correct. We're shifting from being collaborators who shape the process to being supplicants who receive the output. It is a transition from working with a co-intelligence to working with a wizard. Magic gets done, but we don’t always know what to do with the results. This pattern — impressive output, opaque process — becomes even more pronounced with research tasks.

That describes how I obtained from Replit a Virtual Wax Museum of three economists. I made a vague request, and then Replit figured out what I meant by a virtual was museum, decided that the three economists should be Smith, Keynes, and Friedman, decided that it should show some biographical information, bullet points describing their main theories, and include a famous quote from each, etc.

Mollick writes,

We're getting something magical, but we're also becoming the audience rather than the magician, or even the magician's assistant. In the co-intelligence model, we guided, corrected, and collaborated. Increasingly, we prompt, wait, and verify… if we can.

Aaron Renn writes,

One of the most interesting findings to me was the difference in where Gen Z men and women ranked having children as important to them:

Gen Z men who voted for Trump rate having children as the most important thing in their personal definition of success. Gen Z women who voted for Harris ranked having children as the second-least important thing in their personal definition of success.

I think everyone’s previous view would have been that women placed a higher priority on children than men. But for Gen Z, it might be the reverse. Having children would have traditionally been seen as core to female identity, but at least a segment of Gen Z women seems to be rejecting that

Holly Mathnerd writes,

A lot of right-side commentators have been correct when they’ve said that leftism is cluster B at scale. It’s a culture that valorizes narcissism, borderline instability, histrionics, and manipulative cruelty. They’re right about that.

…I doubt the right is capable of the level of evil it would take to defeat the psychopaths, which leads to a very bleak prediction.

But the opposite possibility? That’s no better.

Because out-evil’ing psychopaths is no victory. It is only surrender.

I have many friends on the left who are normies, culturally. I think that Noah Smith and Matt Yglesias and Jonathan Haidt are sane (disclaimer: I don’t know them personally).

There is a large segment of the left that is crazy. They apparently have captured universities and intimidated many people on campus who should know better. They may be in the process of doing the same thing to the Democratic Party, as the rise of Mamdani indicates.

But overall, I think America is still majority normie. I don’t think we need a war to defeat the crazies. And we don’t need to unleash craziness on our side, of which there is plenty.

substacks referenced above:

@

The Psychopath Ratio

@

The Assassination of Charlie Kirk

@

On Working with Wizards

@

Should You Have Kids?