So if Bob, Jim, and Jane are all riding in Bob's car then "y'all have a ride". But if all three of them have their own car, then "all y'all have a ride".
> maximizing prediction accuracy is inherently unbiased
This assumes that you're actually maximizing prediction accuracy, rather than taking the easiest route toward sufficiently high predictive power.
Not-so-hypothetical: you can invest $N and create a profitable model that (unfairly and inaccurately) discriminates directly based upon race, or you can invest $N*M and create a profitable model that does not discriminate on race (regardless of whether its results are racially equitable). Given the choice, a lot of people will choose the $N approach.
Also it should be possible to train the NN model to explain its decision, even in plain language.
I think the "right to explanation" law is pretty sensible, people should have a right to question the algorithmic judgment that impacts their life. It will be somewhat inconvenient for big companies applying ML to make these decisions, but it won't ruin their business.
That largely depends on the depth of the engagement.
Reading Das Kapital should be an exercise in critical thinking, regardless of one's political persuasion. For example, where in the software industry is the labor theory of value a reasonable model? Where is it not reasonable? Why? Answering these questions requires what any reasonable person would call critical thinking.
Conversely, hearing a bunch of passionate lectures about worker's rights movements is considerably less enlightening...
I don't see any problem with passionate lectures about worker's rights movement. In college I also heard passionate lectures about the marvel of free markets, and had to endure them. Different professors have different persuasions and that is all right, they're not trying to fool anyone. It is up to each one to form his or her personal views of the world, and use the information they get from history and philosophy to enlarge these views.
Economics is also the source and home of some of the most bullshit-prone theories and explanations. That's why I ultimately dropped my Economics major. It's full of people that got into the field precisely due to their passion for of their highly politicized opinions. Those people generate elaborate theories, data sets, and models to justify their almost always a priori beliefs -- always under the guise of supposedly rational debate.
I would argue -- as a non-physicist and non-biologist and non-chemist -- that physics or biology or chemistry should take the crown. All explain historical phenomena via models and also extend those models to the present day and even the future, allowing eminently testable predictions. Additionally, the questions these fields pose are typically but not always far less likely to ground out in political preference and are therefore less susceptible (though not invincible) to bullshit.
I was, and still am extremely grateful for my microeconomics prof: he is a contrarian (http://news.nationalpost.com/news/canadian-economist-never-k...) that forced us to reason on a wide class of non-societal examples from first principles whenever possible. As far as being politicized goes, macroecon seems to be a bit more plagued than micro, mod any "homo economicus" fallacies.
Scientists may lack resources and proper incentives, but it's pretty bold to claim that they lack the ability to propose new testable/semi-testable theories...
It's really useful to distinguish technical fields from engineering and especially natural sciences fields once you get past K12. STEM is not actually a very coherent grouping.
Almost all CS programs are relatively flat, but many science programs have longer chains of linear or parallel pre-reqs.
Math programs differ from both because long chains of pre-reqs in Math often correspond to required maturation rather than required knowledge (e.g. the Calc sequence as a pre-req for algebra courses).
Math and CS professors and students complained loudly about the lack of prerequisites. It did great harm to the curriculum. Defending it on the grounds that math and CS don't need long prerequisite chains in the way that other fields do misses the point.
WRT CS, my comment was meant to be descriptive, not prescriptive. FWIW I agree with you, but what I said is largely true of CS curriculum in the US.
I'm not sure whether I agree with you regarding math. I think pre-requisites should reflect required knowledge accurately, and should leave required maturation to the student and his/her instructor. Requiring the calculus sequence prior to courses that contain no calculus strikes me as a silly historical accident, for example.
Technical fields -- especially those closer to technology than to natural science -- can also be extraordinarily ideological. Both in curriculum and (especially!) in culture.
For a perfect example of true ideology in action, look no further than these comments -- from members of a field renowned for epic religiousity over such comparatively inconsequential topics as programming languages, text editors, and software license preferences -- decrying the proneness of other fiends to ideological attachments!
On the contrary, the answer to such a question perfectly illustrates the sort of critical thinking that is taught in a history degree program but not necessarily in a technicaly degree program.
After reading enough about the history of mathematics, it is nearly impossible to arrive at the conclusion that today's mathematicians do not have their own set of ideological beliefs. However, obtaining high scores in a mathematics degree program is easily achievable without seeing these beliefs as ideological ("that's just how it's done" is a profoundly common and profoundly ideological answer to questions of the form "but why not do things this other way" -- especially in upper-level undergraduate mathematics).
1 = "do all of you have a ride (perhaps shared)?"
2 = "do each of you have your own ride?"
So if Bob, Jim, and Jane are all riding in Bob's car then "y'all have a ride". But if all three of them have their own car, then "all y'all have a ride".
ymmv.