I don't find it remotely hard to believe that Ellen Pao's recounting of her experience in the world of venture capital is far from the truth
Can you clarify this? I'm confused because it seems at odds with the rest of your post.
Can you clarify this? I'm confused because it seems at odds with the rest of your post.
For positions and treatment of individuals working in venture capital, however, I don't imagine very many engineers would have much to weigh in on. So what I'm getting at is that I'm pointing out how overloaded the concept of "Silicon Valley" is. Engineering and venture capital are two completely different worlds with the same title slapped on them, and it makes it very frustrating when communicating about things like this. I sympathize with Pao, but the title of the article is literally "This Is How Sexism Works in Silicon Valley."
In terms of "cultural" fit, you're right. I do a lot of work with Hindu Indians and Muslim Pakistanis, because I have an enormous "cultural" overlap with them as a white male from Southern California. Oh, wait, I actually hired them because they were excellent engineers.
Someone might think, "I need to hire an ivy league graduate to do the algorithms work" and feel they are being utilitarian. After all, there must be some value in the incredible price tag of that education.
For me, that's still feeling a bit iffy, but much more reasonable. The more we discriminate, the more discrimination we embrace. At what point is discrimination good and at what point is it bad?
There are clearly protected classes where we are not legally allowed to discriminate. There are classes where we can discriminate, but shouldn't. And there are classes where we must discriminate, otherwise we would just be hiring random people off the street.
And even though I completely agree with everything you wrote, I think there is value in reflecting on what it means to be discriminatory and what focusing on utility might bring us. While I think it is very much beside the point of your post (which I thought was excellent), there is a bit of hubris to gloss over the issue of hiring for utility without examining how this is problematic.
Before you ask -- a different model may be not to hire for utility, but rather to consider hiring as a responsibility you take on. Your responsibility is to help that person be successful. Are you able to do it it? That, in my experience, is a far better view than hiring for utility.
It isn't bias, per se, that is problematic. What creates difficulty are the consequences for poor choices. What we colloquially call "discrimination" is really someone being blind (or uncaring) to the results of their own poor judgement. All poor judgement is harmful, but there are some classes that we single out for special treatment in society.
There are many examples in society where we make "allowances" based on discrimination. We might accept smaller women fire fighters than we do men. We might accept people from certain religions to wear slightly different uniforms. We might allow people with a physical impairment to have an aide for taking a test. By and large, these are choices that are meant to create good consequences, but they are discriminatory nonetheless. Indeed, it is societal bias which determines which allowances are considered to have good consequences and which are not (I expect that some people would question or reject one or two of my points, for example).
Of course, not everyone is good at making judgements. Additionally we sometimes have cultural biases which encourage us to make certain poor judgements. It's important to realise that all poor judgement is harmful. Honestly, it doesn't matter if you make my life a misery because of my skin colour or because you don't like the way I talk. It doesn't even matter if you do it simply because you are unskillful and don't realise (or care) that you are doing it.
But you can't make it illegal to be a jerk. Everybody would be in jail at one time in their life. Because there are some classes of problems that are pervasive in society, we can address the problem there. We can address specific problems with large classes of sufferers. The other potential benefit is that hopefully people will start to realise that their behaviour is undesirable. So if they think, "I can't use sexual innuendo as a way to politically lock out a female rival", they might start realising that politically locking out any rival is not really a beneficial strategy.
I say might. I'm not really expecting that much, but one can live in hope ;-)
Sorry for the book. I feel frustrated in the same way that I imagine you do. I just think that we need to help people steer away from the "thought crime" of bias and start getting them to focus on the "real crime" of harassment. If we complain too much about "bias" (or the allegation of bias), I think we risk people getting stuck on that point and never getting around to discussing how we learn to act better to each other.
Well, it attempts to be utilitarian. However, there are routinely articles on HN, from the perspective of the prospect and the company, about how much of a crapshoot interviewing for coding is.
So fundamentally this debate is about the existence of an objective reality that does not have, at its base, gender or race or other attributes. Most of us who are engineers see a different base and it's pretty condescending to have someone explain to you that you prefer the person who aced the interview to the person who struggled because of some unconscious bias, or that you asked the wrong questions, etc.
When I interview, I have a bank of questions and answers already written out. Each question has a well-defined definition of passing. I decide which question to ask based on the topics that the person lists on their resume. That is, question selection happens from reading the resume, and it happens before the interview. If they say they know javascript, I give them something from the JS pile. If they say they know crypto, I give them something from that pile. I don't want to ambush them with stuff they don't know. The assumption is that if they know the stuff they say they know well, then they can learn what they don't know as needed.
And believe me, there are huge differences in how people do on these questions. For example, in the javascript test, I have some code with vulnerabilities that I ask them to find. I give them the code, explain what it does, explain what I want them to do, and either stay in the room or walk away, depending on what they want. After 20 min, I come back and ask them what they found. A large number of people find nothing. Some find one, but not the biggest one. Some find everything. Some are not able to understand how the code even works. This is for a security role.
If they say they know crypto, then I have a bank of crypto questions. For example, I write down a common authentication and key establishment protocol. Then I ask them what could an attacker do if the last confirmation message was not sent. There are huge differences in outcomes. Some people are confused and have no idea. Others start confidently making stuff up. Others actually identify attacks that would be possible, demonstrating an understanding of cryptographic protocols. None of these differences are based on race, gender, age, etc.
There isn't too much of a middle ground -- most people either succeed or fail. Now, they may have failed because they were having a bad day, or maybe they don't do well at technical interviews, or they might have lied on their resume, but there is an objective reality. This is what people mean when they say the interviews are utilitarian. But I honestly don't think that a large portion of the population is capable of understanding that. They really see everything in terms of identity, so the questions must have been sexist, or I was being unfair to the interviewee.
Now the battle becomes, must everyone else also reject objective reality? Must everyone else throw away the notion of ranking candidates based on their ability to reason abstractly and demonstrate problem solving skills in a live interview, or must we just adopt a notion of "qualification" ? E.g. do we give everyone a job based on their GPA/list of degrees, or is there room for performance based evaluation of individuals? This debate is very important to tech, because there are a lot of incompetent engineers with degrees that really impact a team. There is a big difference in productivity between good engineers and bad engineers. Joel Spolsky is spot on here: https://www.joelonsoftware.com/2006/10/25/the-guerrilla-guid...
In engineering, actual talent matters, and it matters a lot. This isn't like hiring a bus driver. You don't get promoted for having been there for X years, and you don't get hired based on your credentials. Your idea doesn't get adopted because "it's your turn" to have your way. Individual talent matters, and the fact that we may not always detect it accurately is not evidence of the hiring process not being utilitarian or any kind of identity-based discrimination.
Of course I'm not saying there isn't opportunity for bias here as well (mainly in the selection of the question), but you've offset it magnificently by having written down the correct answer and (hopefully) prioritizing getting said answer rather than the manner in which it was conveyed ("I think..", questions rather than statements ("It's x, right?"), "Maybe it's..", etc. since those types of self-effacing comments are often seen and presented as polite forms of communicating for women).
I think most large companies have a formalized system of initial screening -- which has been criticized for relying on keywords and such. It's done by HR. After that, there's a phone screen which is much like an in person interview -- the questions should be pre-written based on the resume and generally they are. In my company I've worked to write up large banks of questions w/answers and distributed them to people doing phone screens for this purpose, and I've used that in my own phone screens.
What people should do is formalize the interview process which helps in so many ways -- a repeatable process with repeatable questions lets you compare people across times and gives a lot of insight. It also helps reduce bias.
I do think that small companies tend to do a worse job that big companies with established HR practices. That's not only true for HR, it's true across the board. Young people who are inexperienced are going to make mistakes. Certainly there is going to be "culture" type bias in start ups and other small companies at least during the screening process. I wouldn't say that's too relevant for "Silicon Valley", as you have these issues in small businesses in every industry.
Traditional barriers to entry, such as having a college degree, tend to be ignored in a small subset of programming jobs (notably Silicon Valley).
Therefore, the hiring process in programming is more utilitarian than many other industries. Your talents and accomplishments matter the most in getting you hired.
Therefore, sexism in tech is exaggerated.
To me, that shows a stunning lack of personal awareness. I've done hiring in tech. I'm an engineer. I'm also a woman. And I know I'm biased against certain groups. I purposefully offset that by making sure there is some sort of measurement (usually a coding exercise) with as objective as possible metrics for judgement (I explicitly state the criteria they'll be judged on so they're aware) that dictates whether they move on to the next stage or not.
If I need to offset my subconscious biases as a woman in tech who spent many of her formative years living in different African countries, than I can only imagine the types of biases a white man who has lived primarily in the suburbs will need to be combatting.
For such a man to be "annoyed" because "anybody who is in a hiring position for specifically engineering knows exactly how utilitarian the hiring process is" is a major part of the problem. You can't correct for biases (often subconscious or quite subtle) you're refusing to acknowledge.
I have a series of technical questions that I ask a candidate specifically meant to isolate important technical abilities. Each of my questions almost obviously highlights one particular problem-solving ability that I expect from a quality engineer, and each question's answer is evaluated to ensure that that one specific criteria being met.
The objectivity by which I evaluate this is subject, from your perspective, because I've alerted you that I'm a white male. However, you know literally nothing about me, the questions I ask, or my primary motive when seeking an employee. I am the director of engineering at my company, and it would be extremely detrimental to my own team if I were to pass up a quality candidate of any race or gender who fits my specific (and purely technical) criteria for a hire.
Your response instead says that I demonstrate a stunning lack of personal awareness, even though not only do you not know anything about me, but I am actively engaging in a conversation about the topic for which I've clearly reflected heavily on. If you want to be less subjective (and quite frankly, racist and sexist) in your own evaluation of an individual you don't know, I suggest that you not make so many assumptions about a person upon discovering that they are a white male.
Consider the high-profile stories of how women get treated in the industry after they are hired.
https://techcrunch.com/2014/03/15/julie-ann-horvath-describe... https://www.susanjfowler.com/blog/2017/2/19/reflecting-on-on...
Even if we assume the hiring process is 100% utilitarian and unbiased (which itself is quite a questionable claim), the tech industry still has a serious problem with the way it treats women once they are hired.
So attempting to argue that the problem comes from the venture capital side and not the engineering side comes across as a little naive.
I've seen so many accusations of sexism in tech by now that I tire of it. When investigated, it seems to always boil down to one of two things:
1. Some guy hit on me and I thought it was gross^Wharassment.
2. I feel a general aura of sexism that I can't pin down or concretely identify but I'm sure it's there, so I've become instantly triggered by mundane things.
All too often the latter feels like women (and male feminists) on a power trip. The feelings of being offended aren't real, it's just a way to convince/force others to do what they want.
I'm thinking of cases like the dongle incident for that. A trivial remark over nothing that led to guys getting fired. But the Damore memo is another good example. Guy points out that lots of scientists think men and women have different interests for genetic reasons, so maybe we should be more willing to discuss that -> outrage, firings, power successfully exercised. Nobody is ever actually offended by science, or even abstract arguments about science and politics, even if they may claim to be. The offence is manufactured to meet an end goal: the punishment of those who do not bow to the power of the ideology.
Please explain this statement further. I am trying to understand how the person who is harassed is the person with the power or on a power trip as you put it. Wouldn't the best way to avoid such a scenario be to avoid the circumstance of making unwanted or unwelcome statements? How in the world can you definitely tell the feelings of being offended are not real?
And how are we supposed to do that?
Depends on if you mean objective, or subjective harassment. Offense could be described as the latter. The power comes from being able to provide the latter, and it being treated as the former.
You can definitely tell in cases where the target of this is a particular individual in a group setting; the double standard in treatment becomes stark after a while.
In general, businesses, regardless of industry, have tended to view complainants as the troublemakers (or at least a legal risk) and their future at the company is affected in one form or another. Anyone that can sue a company is a threat and threats are not typically viewed positively. The individual may achieve a short-term goal (or satisfy their agenda as you point out) but long-term they are not on the winning end of the game they are playing.
Any individual who would use complaints as a career driver, or to drive whatever agenda you are alluding to, is in for a rude awakening. This would apply to those who complain of harassment, whistleblowers', or simply well intentioned employees. We have yet to come up with a way to resolve the issue and protect reputations, career advancement, etc for the parties involved when it comes to unwanted or unwelcome statements that reasonable people can disagree about.
Note: I am not addressing repeat offenders or blatant harassment. I am only addressing situations where a comment was not meant to offend but is offensive to someone but reasonable people differ on if the comment was offensive. Clearly people are different and what offends one does not offend someone else. These are the difficult situations that I am addressing, not someone showing up in a bathrobe at a coworkers hotel room.
The amount of actual time that "long" takes can vary wildly, and assumes that businesses are generally competent at identifying and eliminating "threats." "Long" can sometimes be years of the behavior. Many businesses, even successful ones, are not fully aware of what goes on inside the company.
Even with those issues aside, none of these things you mention will stop you from being the first victim of someone who manipulatively uses complaints about objectionable comments to screw you over. Most places need a pattern of behavior established before they're willing to assign blame to the complaintant, especially if the complaintant is difficult to fire for other reasons (e.g. is a member of a legally protected class or is connected to someone up the corporate food chain).
And yes, I know you're talking about comments and other things that are subjective like that. Those are the ones most likely to be abused, because they can be interpreted in multiple ways.
I'm not going to look it up, but it wouldn't surprise me if science shows that men are more aggressive than women.
It does hurt my feelings when people assume that I'm going to be aggressive/dangerous because I'm a man.
There are many other examples of how science can offend people, but this one affects me the most.
For example, aggression is context bound by the cultural environment. In some cultures aggression would be to talk to an elder without being spoken to first. To argue then that men are biologically more disposed to mechanically perform such aggression is silly, so we would need to disassemble what aggression is and how it plays out in our culture.
If we take violence, we find a similar patterns. Researchers have tried and failed to find any correlation to violent crime and testosterone levels (as long those are within naturally occurring). Similar, a household with two men has equal risk to have a domestic violence happening as a household with two women. Passive aggression is a different topic all by itself, but is clearly defined as aggression.
The science isn't what is hurtful. What hurts is when people hate me for what I am. Being hated for attributes that a person has (but not necessary defines them) is very harmful. Scientific data can sometime cause this and I would say often do when data is taken out of the cultural context in which it is observed. It doesn't make science the culprit, but it does require a more careful discussion by all sides in order for the data to be useful. The Damore memo is an example where such discussion fell apart almost the instant it was created and blame can be pointed at many places except for science.
Now the issue with some studies that can be considered flawed at best being used as a justification would be a problem but that's an entirely different argument here, I feel.
Nope. What they are offended by are assholes using science incorrectly to back seriously sexist arguments that "explain" how they're inferior.
Investigated by whom? Yourself? What criteria are you using, or are you just using your "general aura?"
> The feelings of being offended aren't real
So they are liars?
> Guy points out that lots of scientists think men and women have different interests for genetic reasons
Guy implies that leading to them being less fit for programming without evidence and segues into promoting an alt-right agenda.
> Nobody is ever actually offended by science
People do get offended by poor science used to prop up a political agenda (Global warming, anyone?).
> The offence is manufactured to meet an end goal: the punishment of those who do not bow to the power of the ideology.
So your logical conclusion boils down to a conspiracy theory with the opposing power being a group of lying women on a power-trip. Now why wouldn't people want to enter into a constructive dialogue about gender differences and their effect on the workplace with you to try and improve things?
The people trying to get evolution banned from schools probably are.
Some of the responses I saw to the Damore memo, seemed to boil down to that some things should not be said regardless of truthiness.
Attempts to even look for differences in intelligence or I think personality by race tend to get shouted down rather quickly, because historically they've been used to justify some seriously nasty shit.
The offence is manufactured to meet an end goal: the punishment of those who do not bow to the power of the ideology.
I really don't think it's all that calculated. There are things that people want to believe, or feel it is necessary to believe, and they get upset if you tell them they're wrong (evolution), if you tell them they're only partly right (Damore), or sometimes if you even look like you might be thinking about questioning them (race).