I recently completed Technically Wrong by Sara Wachter-Boettcher. Let me start by saying that I’m glad that Ms. Wachter-Boettcher wrote this book. The tech industry has a lot of issues which need to be brought out into the open and it is definitely a positive development that people such as Ms. Wachter-Boettcher are bringing these issues to the forefront. It really is only recently that people are discussing the continuous erosion of privacy, misogyny in the tech industry, lack of diversity and many other issues. Whilst I would not deny any of these issues, I felt Wachter-Boettcher’s analysis was somewhat lacking and didn’t really get at the realities of working in the tech industry. Wachter-Boettcher cites numerous examples of tech gone wrong, such as a smart scale telling a two year old that he needs to lose weight, FaceBook denying a Native American person an account because it felt that their name was not legitimate, and the abhorrent use of proprietary, black box algorithms to make parole recommendations.
Again, it is definitely a positive development that Wachter-Boettcher and others are writing about these issues, but the alternatives and solutions she proposes seem a bit simplistic. While she doesn’t state this directly, much of the book seems to suggest that all of technology’s woes are caused by the lack of diversity in the tech industry. Specifically that “white guys” from elite universities are running everything. I don’t have an electronic copy of the book, but after about half way through this, I wanted to count the number of times the phrase “white guys” appears in the book. Sometimes this phrase includes Asians, sometimes not.
Is Technology Inherently Prejudiced?
When the YouTube first released their app to allow video uploads in 2013, shortly thereafter, they noticed that many of the videos were upside down. Upon further investigation, the developers realized that the reason why people were uploading videos upside down was because they were left handed and holding the phone in their left hand instead of their right when recording the video. Now you must be thinking that surely the design team would have thought of that, but nope they didn’t. Surely someone on the team was left handed and tried out the software. Maybe and maybe not. At this point, were this to be a more egregious case, such as Google’s Photo Tagging service labeling photos of African Americans as gorillas, or Facebook creating a “Year in Review” movie of a person’s deceased daughter, tech critics usually point to the lack of diversity on software development teams, and that by increasing diversity you bring a wider variety of perspectives on things and hence you don’t have these kinds of problems. Maybe so, but I believe that these problems have more to do with how software is developed today and not the lack of diversity in tech–which is certainly an issue. Let me explain and return to the left-handed example.
The Process Matters
Back in the days before computers, large projects were managed using a technique which became known as waterfall. In a nutshell, projects were conceptualized with very clear goals and milestones and in theory the project flows from one phase to the next. This methodology works well when the outcome is clearly known and changes are not really possible. Consider building a bridge for example. In order to build a bridge, you have to survey the site, take into account all kinds of environmental factors, order the components, most if not all will need to be fabricated. The specifications are not likely to change much once construction is underway. You aren’t going to get halfway through the bridge and decide, “well a tunnel would be nicer”.
This was the way software used to be designed as well. Then along came lean software development and everything changed. You can read about lean here, but lean is a development from the Toyota Production System, and emphasizes minimizing waste. Without going into the whole lean philosophy, software manufacturers discovered that they were spending way too much time and money building features in their software that people weren’t using. (Think Microsoft Office Talking Paper Clip)
The way software is developed now is that instead of building a massive program with every conceivable feature, companies will first build what is known as a Minimum Viable Product or MVP. The MVP is exactly like it sounds… it’s designed to contain the absolute minimum amount of features to be commercially viable. MVPs suck. They are awful. MVPs are often an embarrassment to the developers because they are so incomplete and half baked, but the idea is to get a product into the customers’ hands as quickly as possible, start measuring how people actually use the product, and build newer, better versions as quickly as possible, a process known as rapid iteration.
This process doesn’t work as well for things like cars and buildings where you cannot rapidly iterate, but it works extremely well with software, especially web applications where you can release new features and improvements on a continual basis. Bringing this back to our non-contraversial example. The developers and management probably decided that they wanted an app to upload videos to YouTube. That was the minimum functionality they were looking for. Once users started using it, they realized, “Hey… left handed users hold the phone the opposite way, so their videos will be upside down. We should fix that.” and they did.
Would more diversity help prevent these problems in the MVP? I don’t really think so. I have to believe that YouTube probably had a left handed developer or two on their team and the thought process probably was, let’s see how people actually use this before we spend time and money developing a complex feature that people might never need or want. To a non-developer, this may seem trivial, but there actually is a great deal of additional complexity. In order to guarantee that the video is oriented correctly, the application has to determine the orientation of the device when it took the video OR it has to employ some sort of machine learning on the video to identify the top and bottom.
When you view the many technical issues that highlighted in the book through this lens, they become a lot less insidious and more understandable. Interestingly, Wachter-Boettcher describes an app called Glow which was originally designed as an ovulation calculator to assist women becoming pregnant. However after a while, Glow’s founders (all men) realized that some women were using their app to avoid pregnancy. Wachter-Boettcher blames this on the lack of diversity on the part of Glow’s founders, and asserts that if they had women involved, women would have realized that from the beginning. Maybe so, but I would argue, that in the case of Glow, Wachter-Boettcher unwittingly describes the lean/agile development process and how it actually worked in this case. Glow put out an MVP. It sucked, but was usable and proved the market viability of the product. After a while, Glow’s team learned from their users and adapted their app to what their users actually wanted. Whether there were women or not on Glow’s team, I doubt it would have made a difference because the MVP’s purposes are only to prove the commercial viability of the product and get it out the door as quickly as possible. There are countless examples of companies who come out with an idea, then after getting it in the users’ hands, realize that users actually wanted something completely different, pivot and become quite successful. Paypal is one such example. PayPal was originally a way for PalmPilot users to send money to each other.
To be clear, I am not advocating against increased diversity in tech, but my hypothesis here is that the problems Wachter-Boettcher cites are caused more by the lean/agile methodology than inherent biases and the lack of diversity in high tech. I also am not advocating against the lean methodology. It has dramatically changed how software is developed for the better, but it has resulted in some pretty spectacular failures as well.
Edge Cases are not Always Racist
Wachter-Boettcher addressed the notion of edge cases. An edge case is an exceptional situation which can cause a system not to work as intended. She uses the example of companies not including multiple gender options. Software companies would assert that transgender individuals make up an extremely small percentage of the population (and user base) and hence it isn’t a priority.
Let’s consider another situation. I am Jewish and Orthodox. This means that I don’t turn things on or off during the Sabbath (Shabbat). As an ever increasing number of devices around the home are becoming “smart” they also are becoming problematic to Orthodox Jews. The Nest thermostat for instance, lights up when you walk by. Does this make Nest anti-semetic because they are not accommodating Orthodox Jews in their products? Should Nest have an Orthodox member of their engineering team so that future products will be Sabbath friendly? (As an aside, many appliance manufacturers actually do have Sabbath Modes because Orthodox Jews represent a significant market for high end kitchen appliances) This case certainly is an edge case (or stress case as Wachter-Boettcher refers to them), but is this an indication of malicious intent? I would argue no, but rather simple economics.
A lot of tech products are developed by small teams. Wachter-Boettcher would note that there often isn’t much diversity on these, but the main thing to note is that these teams are small. They are often under a lot of pressure to produce workable products under tight deadlines, so they have to prioritize. Are they going to build a feature that 0.001% of their users care about, or are they going to build a feature that 95% of their users care about? It seems clear to me that no matter how well intentioned or diverse a team is, they are going to prioritize the things which maximize their chances of building a commercially successful product. Hence, Nest has no Sabbath mode, and my iGrill temperature gauge doesn’t have an option for me to hide the pork. (Not that I really care about that)
What about Free Speech on the Internet?
Wachter-Boettcher addresses how many online fora have degenerated into caldrons of hate, abuse, racism and sexism. She highlights how a popular actress was harassed on Twitter, proliferation of child pornography and other vile speech on Reddit and more. Wachter-Boettcher’s position is that these sites (and others) need to do a better job of policing their sites to remove offensive content so that these sites can be safe for everyone. For the purposes of this argument, let us distinguish between illegal speech such as specific threats of violence, and speech that is merely offensive. The bottom line here is that I do not believe that it is possible to police offensive speech without expressing a view point of some sort and censorship. Let me elaborate. Liberal organizations such as the ACLU and others, actively work to protect anyone’s right to express themselves, no matter how vile or how much they personally disagreed with the speech itself. As an example, in 1978, the ACLU fought for neo-nazi’s right to demonstrate in a largely Jewish area of Chicago and even more recently, in 2012, the same organization fought for the kkk’s right to distribute flyers in Missouri. I’m reasonably confident that nobody at the ACLU agrees with the kkk or neo-nazi’s opinions, however they will fight for their right to express these opinions without government interference or censorship because they feel that free speech is an absolute right.
Many people in the US Military feel the same way about flag burning. They find it extremely distasteful, but will put their lives on the line to defend your right to do so. Bringing this back to Facebook, Reddit and Twitter, this standard does not seem to apply. Again, I am not talking about abuse, but if a site purports to be a site that is open to everyone for a free and open exchange of ideas, then it has to be open to distasteful ideas as well. If a site is claiming to be a safe space for all its users, then it has to censor some users, because inevitably some users will find speech offensive that others will not. Which brings us back to the original question, how does a site limit speech without expressing a point of view? I would challenge the reader to pick a controversial topic (Yankees vs Red Sox for instance) and argue one side in such a manner that is guaranteed not to offend the other side. I think you’ll find it is quite difficult if not impossible.
But why do these sites exist in the first place?
When people criticize sites like Facebook, Google and others, they seem to forget that these sites exist to make money. Despite Mr. Zuckerberg’s platitudes about bringing the world together, Facebook exists solely because it is generating billions of advertising dollars in revenue. If Facebook did not do that, it would cease to exist at some point. It is not a “community forum dedicated to the free and open exchange of ideas” but rather a sophisticated platform designed to gather data about its users and serve them targeted ads. Full stop. If you view their actions through this lens, Facebook (and many the other companies’) behavior makes a lot more sense.
What can be done about the lack of diversity in tech?
Without question, this is a tough problem and I don’t have any easy answers. Wachter-Boettcher correctly points out that in discussing this issue most hiring managers in tech would point to the pipeline as the issue. Specifically that there aren’t enough minorities and women with the requisite skills. Wachter-Boettcher is very dismissive of this issue, but it is very real. Don’t believe me? Go to LinkedIn and do a search for “Java developer” or “Systems Architect” and see what comes up. Go to pretty much any tech conference and take a look at the lines for the washrooms and you’ll clearly see the gender imbalance.
The issue as I see it is that the lack of diversity occurs early in the pipeline. Wachter-Boettcher quotes a statistic stating that something like 16% of all computer science graduates are female. I don’t recall the stats on minorities, but they were also very unbalanced. This echoes my personal experience. My graduating class at the University of Arizona Computer Science Dept. had three women in it out of 60 or 70 total. If only 16% of the graduating computer science classes are women, how can tech companies increase the number of female developers?
I would therefore argue that tech companies cannot be faulted for not hiring women and minorities if enough women and minorities aren’t obtaining the requisite skills to work in the tech industry. But why is that the case?
Again, I don’t have a good answer for this, but I think the answer lies in making technology more appealing to everyone at an earlier age. The important thing here is to distinguish between merely using information technology, which practically everyone does, and actually building high tech, which is not so popular.
When I was a wee lad, the archetypal computer programmer was a nerdy guy with thick glasses and a pocket protector. It wasn’t “cool” to be interested in computers. As a result, there was a kind of self selection whereby only the “nerds” played with computers, and ultimately studied computer science. I graduated high school in 1996, and I would strongly suspect that a lot more people in the marching band had email addresses than on the football team.
I don’t think that technology has the same stigma today as it did when I was in school, but at the same time, from my own observations, young people today are interested in using technology, but not necessarily interested in learning how it works so that they can build it themselves. This is a serious problem.
Put it in another context. In the US, most people drive cars or generally familiar with how to operate one. With that said, comparatively few people know how a car actually works. The people who are really interested in cars, learn how they work, how to repair or modify them and go on to become automotive engineers or mechanics. The same is true of technology.
I did a VERY unscientific survey on LinkedIn where I asked the question what made you get into tech, and I got some really interesting responses. I noticed that most of the women who responded said something about a teacher or other role model encouraging them, usually when they were in high school or early college, whereas most of the men kind of said they knew from an early age. I will leave this question to social scientists and educators, but I would be interested in researching this question further.
What about Mysogeny in Tech?
This has to stop. Women in tech need to be treated with the same respect as men and the reports of how some CEOs and other senior staff at various companies have behaved are really disgusting. Seriously, this has to stop. When people treat women badly and disrespectfully in the workspace, aside from being morally wrong, it drives women out of the industry. Organizations absolutely must not tolerate bad behavior regardless of whether the person is a “high performer” or not.
With that said, there are jerks in every profession and of course some professions attract more jerks than others, but I still think tech is a great industry to be in largely because of the meritocratic nature of the industry.
Meritocracy? Myth or Reality?
Wachter-Boettcher is highly critical of the concept of meritocracy and dedicates an entire chapter to criticizing the mindset. In chapter 9, Wachter-Boettcher discusses how she perceives that the meritocracy has broken down, if it ever existed. However, I view things a little differently… Wachter-Boettcher is correct in noting that many founders of companies have similar backgrounds and social connections, and she almost exclusively focuses on this. However, what she seemed to miss is that this is not the case for the technical leaders and workers in the industry.
Here’s the thing… The tech industry is one of the very few industries in which a person can do very well for themselves without an academic pedigree. In fact, unlike many other professions, there really are no gatekeepers to the profession, no licensing boards, nothing really to keep you from landing a good job. All you have to do is develop your skills. Furthermore, there are TONS of free or near free resources available for you to do so! But… you have to be self motivated, passionate, and determined. Tech isn’t easy. But I’ve known people who have been able to raise themselves up from poverty or other bad situations to have successful careers in tech. Furthermore, since there are a lot of people who rose up through the ranks in the same manner, they are willing to give people with unorthodox backgrounds a chance.
Conclusion: What can we do?
In Technically Wrong, Wachter-Boettcher does a good job of identifying issue in tech, but from my perspective her recommendations on how to rectify them need further examination. However, her book is a good perspective and I would definitely recommend reading it. Clearly these issues are extremely complex, and there are no easy solutions for any of them. Awareness is a good first step.