Diversity of thought is a pretty pale substitute for actual diversity. As someone flippantly pointed out elsewhere in this thread, melanin is not midochlorians. BUT melanin is statistically correlated with lower median household income and accumulated wealth, with being incarcerated or having an incarcerated family member, with single parenthood, and with being misidentified by AI.
When I look at the above group I don't actually see that much diversity of thought. I see a group of people who are likely not primary caregivers for children or elderly parents. I see a group of people who is more likely to be promoted to management if they choose that career path. I see a group of people who were affluent enough to have access to personal computers and professional development related to personal computing in the 1990s, before this was super common. I can't list the CV of everyone on this list, but it seems to me that this group also skews toward consultancy shops as opposed to product shops. These things all inflect the base assumptions that are baked into the manifesto. The question of whether agile software development constitutes good development praxis is inseparable from the diversity of the philosophy's founding group.
For example, if you have a lot of spare time to devote to professional development outside of work hours, you might press for more (or all) of your hours at work to go to pair programming. Pair programming in itself isn't a bad thing, but if it's all you're doing, you might be eliminating time that someone would ordinarily spend studying design patterns, or experimenting with refactoring methods on their own to get practical experience. If you think that all of that sort of professional development should take place outside of work hours, you're implicitly stating something about the kind of person who ought to be able to work in software. This isn't just an arguably bad moral stance, it leads to worse software.
If you don't have women on your team from the beginning, you're less likely to develop anti-harassment and safety tools. See also: Twitter. I bet someone with incarcerated family members would probably have something to say about the development of prison e-readers that are used to restrict access to books for incarcerated people. Likewise, they might have valuable insight into algorithms that predict recidivism rates that are based on biased and incomplete datasets that target residents of low income and black neighborhoods. Having trans people on your team is great for privacy and data collection standards, because they know to question the data collection fields that you might think are a no brainer, but are actually unnecessary for you to store!
I take your point that these seventeen men are not identical. But their collected viewpoints fall so radically short of representing and addressing the needs of a growing and shifting population. Software is made for people, by people. Right now, a lot of people's valuable insights are dismissed by virtue of being excluded from the room. These seventeen men might have come to a consensus over a long weekend. But they stopped there. They didn't solicit amendments or context reads from their diverse colleagues. In 2001, it was likely that they didn't have many diverse colleagues. That had important ripple effects that are still manifesting. This should be instructive for how we compose our teams and build our products.
Diversity of thought is a pretty pale substitute for actual diversity. As someone flippantly pointed out elsewhere in this thread, melanin is not midochlorians. BUT melanin is statistically correlated with lower median household income and accumulated wealth, with being incarcerated or having an incarcerated family member, with single parenthood, and with being misidentified by AI.
When I look at the above group I don't actually see that much diversity of thought. I see a group of people who are likely not primary caregivers for children or elderly parents. I see a group of people who is more likely to be promoted to management if they choose that career path. I see a group of people who were affluent enough to have access to personal computers and professional development related to personal computing in the 1990s, before this was super common. I can't list the CV of everyone on this list, but it seems to me that this group also skews toward consultancy shops as opposed to product shops. These things all inflect the base assumptions that are baked into the manifesto. The question of whether agile software development constitutes good development praxis is inseparable from the diversity of the philosophy's founding group.
For example, if you have a lot of spare time to devote to professional development outside of work hours, you might press for more (or all) of your hours at work to go to pair programming. Pair programming in itself isn't a bad thing, but if it's all you're doing, you might be eliminating time that someone would ordinarily spend studying design patterns, or experimenting with refactoring methods on their own to get practical experience. If you think that all of that sort of professional development should take place outside of work hours, you're implicitly stating something about the kind of person who ought to be able to work in software. This isn't just an arguably bad moral stance, it leads to worse software.
If you don't have women on your team from the beginning, you're less likely to develop anti-harassment and safety tools. See also: Twitter. I bet someone with incarcerated family members would probably have something to say about the development of prison e-readers that are used to restrict access to books for incarcerated people. Likewise, they might have valuable insight into algorithms that predict recidivism rates that are based on biased and incomplete datasets that target residents of low income and black neighborhoods. Having trans people on your team is great for privacy and data collection standards, because they know to question the data collection fields that you might think are a no brainer, but are actually unnecessary for you to store!
I take your point that these seventeen men are not identical. But their collected viewpoints fall so radically short of representing and addressing the needs of a growing and shifting population. Software is made for people, by people. Right now, a lot of people's valuable insights are dismissed by virtue of being excluded from the room. These seventeen men might have come to a consensus over a long weekend. But they stopped there. They didn't solicit amendments or context reads from their diverse colleagues. In 2001, it was likely that they didn't have many diverse colleagues. That had important ripple effects that are still manifesting. This should be instructive for how we compose our teams and build our products.
THANK YOU Jess. Yes. I really appreciate you lifting up specific ways lack of representation has led to extremely harmful products and practices.
That's an awesome reply. Thanks for expressing all that so eloquently.