As a theoretical example, consider how would you pick up Megan McArdle - she writes, sounds and looks a lot like my past girlfriends, and Suderman looks and sounds broadly like the same kind of guy I am. This just a hunch, though.

(..)

On speed dating events in Birmingham, there was a non-fat, intelligent, friendly, considerate 15-20% always.

Just a hunch but I suspect Megan McArdle would not be doing speed dating.

Autonomy means people can decide to be different from each other, and thus be really cautious with generalizations and stereotypes

Except the generalizations are frequently correct and have enormous predictive power.

perhaps, cultural ones are still okay, because socialization is a powerful thing, but gender is not a culture.

Why? Yes, socialization is powerful, but so is genetics, including the difference between XX and XY. In particular the SRY gene has much more influence than a typical gene.

Second, and more important, the ends not means stuff means not seeing sex as a prize to be won by an active, driven men and women just passively hand it out as a reward for the effort, but as an mutually initiated, mutually desired interaction between two autonomous beings with their own desires.

You see to be confusing is and ought there. However, you think sex ought to be obtained, being active and driven (among other things) makes a man more likely to get it. Whether, you consider the women's behavior here "passive" or "actively seeking driven men" is irrelevant, and probably doesn't correspond to any actual distinction in reality.

Objectification is not necessarily sexual and it is really an old idea, not some later day SJW fashion. It is treating people as means. Marx argued that in a 19. century factory the proletarian is objectified into being treated like a human machine.

So you're saying its not just SJW because it was also used by their leftist predecessors?

An object is simply something that does not have own goals, it is the object of desire, or the tool to achieve other desires with, of other people. If you understand what being a person, what personhood means, well, objectification is just a denial of it.

If you mean that humans are game-theoretic agents, I agree. However, I don't see how "therefore we can't or shouldn't apply probability theory to them" follows.

I would say objectification is largely a modern phenomenon, a phenomenon in an age where machines and processes are so predominant that we tend to see people like them, too, and the essence of personhood - intellect and will - gets ignored.

Doesn't this seem to contradict your earlier claim that anti-objectification was responsible for the abolition of slavery?

The intelligent asshole will perhaps present a bogus physical theory to gain status - but the arguments will be about a commonly understood, verifiable thing outside himself. But a social theory will not be about a thing, it will be essentially about himself, something only he really knows and we can just guess.

Well, in this case the social theory in question is indeed about a verifiable thing outside the person, namely the dynamics of human romantic interaction.

Interestingly, Rothbard and Austrian Economics have something interesting to say here, the limitations of empiricism about people's behavior. You need repeatable experiments. But if you repeat it with different people, that is not really valid because people are far, far too diverse - remember, autonomy.

Quote please. I'm guessing you're badly misinterpreting what they wrote. Probably something about how since people respond to incentives, empirically observed behavior will change when the incentives change. Something like a proto-version of Goodhart's law. This is not the same thing as the claim that the laws of probability don't apply to humans, which is the claim you seem to be making.

If I repeat a behavior experiment with two different groups of people and get something like 62% an 65% do X then of course that means something, but it is not, strictly speaking, the repetition of the experiment.

If you mean there is a lot of variance among humans, I agree. However, you seem to be arguing that we should worship and/or ignore this variance rather then studying it.