I suggest you reread #3 and note the use of language like "suggest", "consistent with" and "perceived". Authors use language like this because its really hard to make definitive statements in this area. In addition, its really hard to quantify the size of the effect and how much is due to nature vs. nurture.
Furthermore, I'm relatively certain the authors are not experts on software engineering. As you have likely seen at this point, you could also claim that higher empathy actually should incline people to the job (see, e.g. https://medium.com/@yonatanzunger/so-about-this-googlers-man...)
See citation above. I'm saying its much too strong a claim to go from this paper to the conclusion that women are biologically less inclined to engineering.
EDIT: To be consistent with my original statement (see child)
You made a blanket statement about discussing offensive things and I wanted to point out there are some limits, especially at the workplace.
In this particular case, I feel that leadership and everyone who disagreed with the memo should have at least stated explicitly (if briefly) what the flaws in the reasoning were, as I tried to do at the beginning of this thread.
Obviously, my explanation was not sufficient to convince you and you are still holding on to the belief that women are biologically ill-suited to be engineers. But at least now I have tried to make it clear to you where I think you are wrong. Thats pretty much all you can accomplish on some issues.
Well, saying things that are clearly wrong and offensive to large numbers of people is typically looked down upon amongst educated people and highly discouraged at the workplace.
For example, think of other stereotypes about other groups (blacks, hispanics, jews) and think about what the response would be if an employee was publishing 10 page manifestos with loosely connected scientific arguments in support of these views.
Maybe. I don't think Sundar Pichai or Danielle Brown did though, they just said "this is not OK" and "code of conduct". Perhaps they felt they didn't need to engage with this kind of argument, but that mentality is exactly what has a lot of people upset.
You're right, my one sentence is just trying to explain the basic point to people who may not get it. Its not trying to lay out a complete thesis in support. I'm not sure if thats a necessary exercise, especially for the CEO.
A lot of people on "our side" did a bad job of simply refuting the argument presented. Its straightforward to do in one sentence:
"While there may be scientific evidence of differences between men and women, using these differences to conclude that women are biologically less inclined to engineering is a gross leap in reasoning that is not at all supported by the facts."
Then you can go on and explain why for historical and societal reasons putting forth this weak theory is highly offensive and damaging to a large group of people.
Of course, many feel that you shouldn't _have_ to explain this to people, but unfortunately in todays environment, you do!
I agree that the document blurs the line between preference/aptitude and is not totally clear which is one of the reasons overall it is a mess.
Nonetheless, many of the statements in "Personality differences", "Men's higher drive for status", and "Non-discriminatory ways to reduce the gender gap", especially the stuff about women being more prone to anxiety, liking part-time work, caring more about people than things, etc. speak strongly to aptitude for engineering, given the context of the document.
EDIT: Also, I reject the implication that the claim "women have less of a preference for engineering" is not in and of itself a harmful stereotype.
Many methods in machine learning and statistics benefit from two types of theory:
1) Optimization theory - which says, if I repeat this iterative method N times I will find a (nearly) globally optimal solution
2) Statistical theory - which says, if I observe this process N times I can accurately estimate a population quantity with high probability
Deep learning does not benefit from the same theoretical guarantees.
For the most part the response from the community is "but it works really well!" which is a fair and valid response especially since what most practitioners care about is predictive accuracy.
Personally, I find applying neural networks extremely annoying at times due to the amount of twiddling of hyperparameters, slow convergence, etc.
Wow this guy really needs to take it down a few notches. Yes, PG's articles present a magical fantasy world dominated by the masterful hacker and visionary entrepreneur. They are inspiring and fun to read. And they may have some truth to them.
Attacking these essays as an oversimplification while presenting a caricature of the introvert and disconnected programmer is borderline ridiculous.
Theyre writing long term contracts for power production which provide a fixed price for the resource developers and by consequence provide them with a fixed cost.
Typically, these types of contracts, providing price certainty, are required to build develop new renewable assets.
I suggest you reread #3 and note the use of language like "suggest", "consistent with" and "perceived". Authors use language like this because its really hard to make definitive statements in this area. In addition, its really hard to quantify the size of the effect and how much is due to nature vs. nurture.
Furthermore, I'm relatively certain the authors are not experts on software engineering. As you have likely seen at this point, you could also claim that higher empathy actually should incline people to the job (see, e.g. https://medium.com/@yonatanzunger/so-about-this-googlers-man...)