So the claim is that playing with computers at a young age is vital to the development of programming skill, just as the Hasbro game "Operation" is for physicians.

See: "extraordinary claims...".

No, again, you're creating a strawman. I never said it was vital; just that it helps a lot. And I see no reason such a claim would require a similar effect to be present in every industry.

In fact, the evidence supporting the idea that programming is special in this regard is precisely that you don't hear doctors talk about how they had spent their childhood years endlessly playing Operation, while meanwhile you do hear many great hackers say they spent their whole childhoods playing with computers (or, when unavailable, then the nearest facsimile: taking apart electronics, playing with Game Genie codes, building erector sets...)

It's hard to tell whether you're missing my point or agreeing with it. The idea that computer science is unique among the sciences and professions --- including detail, abstraction, and symbol-intensive fields like aerospace engineering and forensic accounting --- is itself an extraordinary claim.

It's also a counterintuitive claim, given the reality of what software development work actually is. A plurality of professional developers do little more than plumb connections from SQL database tables they understand poorly to front-end frameworks they didn't write. That's the practice that requires exposure from preschool, like a competitive concert violinist?

No. Not buying it. Sorry.

Indeed, most tech industry jobs don't require the sort of deep hacking abilities you can only learn young. However, the hiring process puts high value on having learned to hack young, and the social scene tends to exclude people who didn't have lifelong exposure to hacker culture.

Changing both these things would improve the gender balance.

I'm glad to see you writing that, and thanks. Lemme just add that I think I might be able to surprise you about aptitude to and development of "deep hacking abilities". I don't know about robotics, but C/ASM RTOS-style systems programming? I think you can quickly develop that from people with no prior exposure. I built a hiring process on that hypothesis, and months after leaving the firm we used it at, it apparently continues to be working very well.

I think if you were able to somehow come up with a compelling way to get kids to play forensic accountant (perhaps a new J.K. Rowling series?), you would absolutely see in 20 years that the kids who did were significantly more likely to go on to become great forensic accountants. Same with aerospace engineering. The special thing that distinguishes programming from those skills is that programming skills do have many analogs in the world of childhood playthings and can thus give certain children a preparatory advantage over others when they all reach college together.

Actually, aerospace engineering might have a bit of that. Have there been studies to see if children with the opportunity and encouragement to design, build, and fly their own model aircraft might perhaps be more likely to grow up to be successful aerospace engineers than children who aren't exposed to such things?

Your intuition that the world's best forensic accounting game would produce a generation of the world's best forensics accountants might be useful, but it is not evidence. And the requirement here is for extraordinary evidence.

What do you see as my corresponding extraordinary claim?

Once again: the idea that computer science is unique among virtually all the professions in that performance at the median level of competence requires, like a concert violinist, exposure at the earliest possible age. This despite the fact that most of what programmers do is simply plumbing and diagnostics.

> performance at the median level of competence requires [...] exposure at the earliest possible age.

I never claimed that, though. I think my actual claims were pretty mundane:

1. Children who play with computers have an advantage when, later in life, they're learning programming.

2. Young girls, in many places, are not encouraged to play with computers the way that young boys are.

It's funny that you keep mentioning highly quantitative fields like aerospace and actuary, and then try and contrast them with computing.

http://www.ae.utexas.edu/news/students/683-more-women-ase

Undermines your point a bit.

This is sleight of hand. What you're doing here is crossing the streams between your conversation with me, which is about intrinsic male advantage in math ability, with my conversation here, which is about the supposed advantage men get from being exposed to computers at an earlier age.

The point of citing fields in your discussion with me is that there are quantitatively intensive fields which have significantly greater female participation than computer science, despite the dubious idea that computer science as it is actually practices is quantitatively more demanding than, say, the demands on corporate lawyers to structure debt and equity financing.

The point of citing fields in this particular discussion is that there are demanding technical fields in which practitioners generally get little to no exposure to the work until they approach the age at which professional software developers enter the industry.

Did you miss the distinction? I sometimes go "on tilt" in discussions and misfire like this too.