Well incorrectly according to whom? People are well aware when something is wrong. They just love the money. A training course won't solve that problem. Brainwashing one will though.
Lots of clever people unintentionally trained deep networks that unfairly discriminated against people of various races or from certain zip codes. Sometimes you aren't aware of all the implications of the decisions you might make and a course that informs you of such pitfalls can be helpful. Professional ethics courses at the University level aren't there to convince you discrimination based on gender, race etc is wrong, but to make you aware of when that may be occurring.
I don't know. From what I remember of the flame war the problem was that they fairly discriminated against various races and zip codes. So actually the algorithms had to be made less accurate to fit the legal expectations.
But once again the problem was in the customers. They should have said optimize for a, while keeping discrimination in check, not just optimize for a
So, do you think calculus, physics, type theory, etc are all things that should be taught and studied, yet ethics is not? Do you think that unlike the former subjects, the latter is just something that everyone knows?
The fundamental difference is that calculus, physics, type theory etc are all studies of what is. The goal is simply to describe and predict the world.
Ethics is the study of what one ought to do. It purports to lift some values over other and define morality, meanings, and legitimate goals.
These two - positive versus normative - are fundamentally different types of discussion and knowledge.
Just consider: A physics course in Nazi Germany or Communist Russia would likely teach the same thing as a physics course in Stanford. An ethics course absolutely would not. Consider why and what that means.
> Just consider: A physics course in Nazi Germany or Communist Russia would likely teach the same thing as a physics course in Stanford. An ethics course absolutely would not. Consider why and what that means.
Deutsche Physik rejected relativity and quantum mechanics, aka “Jewish physics”:
Philosophy, including ethics, is foundational to everything we do, and unfamiliarity with the rules of the normative allows the unscrupulous to corrupt the positive.
You're right. I was thinking of things like rocket physics (the US gained many great engineers and physicists from German and to some degree Russia). Example would have been better if I specified "rocket physics" or "aircraft engineering". Let's proceed on that basis.
"unfamiliarity with the rules of the normative allows the unscrupulous to corrupt the positive. "
Which rules of the positive are you talking about? Your rules? My rules? Trump's rules? Hitler's rules? Confucius' rules? Jesus' rules? Mohammed's rules?
The point is that the normative doesn't have rules in the same way as the positive. You can do a science experiment to objectively show Lysenkoism is wrong. You can't do a science experiment to objectively show that slavery is wrong.
Given that there are so many systems of rules of the positive, there's no way you can teach one as the rules the way you can teach physics.
My fear with such courses is that they just end up as moral propaganda for whoever is in power - empowering the powerful.
> Which rules of the positive are you talking about? Your rules? My rules? Trump's rules? Hitler's rules? Confucius' rules? Jesus' rules? Mohammed's rules?
I'm talking about the ones that explain how to think about concepts, how point-of-view and experience affect what we perceive, how discourse can be used to further or detract from truth, etc. Basically, epistemology. Otherwise, how do you even presume to tell me that your physics is right? Because rockets fly? I think they fly because if you put fuel and make an offering of electricity to the gods of the ether, they will send it upwards—on what basis do you convince me, when I can rephrase everything you say to me as "the gods of the ether will it so"?
What has happened is that the Western world has created a shared epistemology and has done a very good job of laying it down and universally teaching it. So most of the time, we don't need to worry about right and wrong because the decision has been made for us long ago (and what does that say about us?)
Now, though, new questions are coming up—ethics in software engineering, bias in machine learning, etc.— where the universal model hasn't yet caught up and been agreed on, or where it is being challenged. And if we are not familiar with the process by which such things are agreed on, we are essentially letting other people make the decisions for us.
I totally understand the reluctance to empower the powerful, but I think education doesn't have to be brainwashing.
> The fundamental difference is that calculus, physics, type theory etc are all studies of what is. ... Ethics is the study of what one ought to do. ... Just consider: A physics course in Nazi Germany ... would likely teach the same thing as a physics course in Stanford
This is a superficial understanding. Once you start digging deeply, the difference between the hard and moral sciences is more tenuous, and less "fundamental". Sure, "calculus, physics" etc largely attempt to describe what "is", but these fields are motivated by the belief that nature "ought" to make sense. There would be no calculus or physics without a desire to make sense of the world, an "ought".
And "ought" does not just motivate hard science, it's an integral part of it. When following the scientific method, the first thing you do is form a hypothesis. A hypothesis, but its very definition, is not necessarily true, i.e., it's an "ought" not an "is". Scientists therefore have to embrace and believe in "normative" knowledge to advance just like other endeavors. The types of hypothesis and endeavors hard scientists engage in are subject to the prevailing social norms.
To reference your example, Nazi Germany did all sorts of ghastly experiments on twins and other prisoners, all in the name of science. Something that would not be done elsewhere. Even today, research on areas like stem cells and global warming are largely influenced by social norms and "ought".
Accordingly, while I agree there is a fundamental difference between "is" and "ought", that difference does not so clearly differentiate hard and soft sciences.
I think that many philosophers have gradually come around to the idea that the fact/value distinction is illusory. This is particularly apparent in the case of so-called "thick" ethical concepts (words like "cruel" or "courageous"), which do not seem to belong exclusively on either side.
A good software written to spec is by definition ethical. It is unethical to write bad software, not evil one. But if you write bad software - nobody should hire you anyway.
> A good software written to spec is by definition ethical.
Even if the spec specifies unethical behavior? Sounds like passing the buck.
Or what if the spec itself isn't well defined, and leaves many undefined states and holes that the developer has to fill in themselves?
If a lawyer submitted a brilliant and conforming but racist or unethical argument to a court, the lawyer can still be found to be unethical. Just because you're following the rules, doesn't mean that you're doing the ethical thing.
Even if the spec specifies whatever unethical behavior means. That is problem for the guy using the software not the one making it.
Get the job done, get the money, get out.
If you want to waste precious time of your life wondering whether something is wrong - be my guest. But don't impose it on people with better things to do.
