Diversity is not just about fairness

While on holiday I posted a link to this BBC article about what the internet might be like if it was run by women. This generated some unexpectedly hostile responses which I was not physically or mentally equipped to deal with. I shall now attempt to do so.

I've decided the best way to proceed is to set forward exactly what it is I like about the article and why I think it's important. I will mention some of the criticism but I'm not going to do a detailed breakdown of this. Nor will I get into an argument about hostility, save the following single point : anytime someone mentions "spoon feeding" or "kid gloves", let alone "nappies", for heaven's sake*... in my book it's a huge red flag that someone is actively trying to give themselves an excuse to be hostile. As with people who respond to articles only to say "well I knew that already", there is just no need for this. Better by far to discuss the arguments, which are a good deal more interesting than getting into a row. There's plenty of better things to row about anyway.

* Frankly, coming from one of the best people I know on social media, this one stung. And I don't understand why my saying "I like this article" should lead to a response like this.

There's one other preliminary point. I've mentioned before many times that interpretation is subjective, partly in response to raw numerical data but also when it comes to narrative fiction. Unless an author is absolutely explicit, you can read events described as good or bad to suit your fancy. The importance of the "eye of the beholder" applies to factual articles too. So here is my interpretive reading.

I believe diversity is important for more pragmatic reasons than simple judicial fairness. It matters, I think, because different perspectives are inherently valuable : nobody has all the answers (see John Stuart Mill), and the final collective response can be far more than the sum of its parts. For example, the way the CIA incorrectly assessed the threat posed by Bin Laden, the difference in masculine and feminine perspectives in philosophy, of course also all of politics, and not least in establishing a scientific consensus... to me the notion of the pragmatic importance of a diverse range of viewpoints seems to have overwhelming support. Now the details of this are frightful*, but the basic premise - the potential for a mix of viewpoints engaged in good-faith debate to result in a better output - is to be almost self-evident. I'm aware that I have a strong ideological stake here, but I do not think this is without foundation.

* The mantra "only connect !" is clearly flawed : an example I've used before is if you take some random vegans and fox hunting enthusiasts and lock them in a room together, you shouldn't expect a productive outcome. No, managing diversity to get a productive result is hugely complicated. My hypothesis is more that diversity is necessary for a more accurate consensus, but definitely not that it is a sufficient condition. Assessing sufficiency is a huge topic and I'll not attempt anything so grand.

So, if we put more women in charge of the internet, what would happen ?

Note that this is the question in the headline. Not "would it be better ?" and certainly not "how could we make it better ?". Simply what would happen. To be fair, the article does naturally allude to these points. It certainly has a strong "this would have helped" vibe to it. But asking "what would happen" in no way at all can be said to equate to "would this be better". Those are fundamentally different question. We could ask what would have happened if Napoleon's empire had never fallen, or if Nazi Germany had developed the atomic bomb, or if the Earth was five million miles further away from the Sun or orbiting a giant cat... Asking the question in no way means the author is inherently suggesting it's a good idea, even if there is often a thinly-veiled aspect of this to the article.

My working hypothesis is that there might be some weak natural mental differences between men and women, but these are so overwhelmed by cultural factors that they can (and probably should) be neglected. The world is complex enough that even asking the question "which gender is more intelligent ?" is truly meaningless; differences there might be, but intelligence is a slippery concept indeed, and highly context-dependent.

So does this mean I think gender is irrelevant ? Alas no, because gender (at the very least the social aspect) is also a cultural construct. Men and women face different sorts of prejudice, so even if the innate differences can be neglected, the cultural factors certainly can't. Given these radically different, unfair social pressures to conform to different norms, it would be surprising indeed if this didn't result in some very different perspectives in different areas. As evidence, I cite every single woman I've ever met.

This to me is what's interesting. In all honesty I'm not terribly concerned here about the precise details of how we'd restructure the internet for the betterment of mankind, for the present, I'm far more interested in those different male and female perspectives. That's what I think the article does well. In terms of of analysis of how to make a better internet, I will cheerfully concede that it's basically shite -  but that's not what I think the intent of the piece is.

What then does the article suggest might happen ?

A more varied team might also result in an internet that is more international in its outlook with a greater sensitivity and awareness of non-Western cultures. This would avoid embarrassing "foot-in-mouth" situations, such as in 2016 when the Indian Supreme Court ordered Google, Yahoo, and Microsoft Bing to remove ads offering ultrasound and prenatal sex-testing – services illegal in a country with one of most unbalanced gender ratios and highest female infanticide rates in the world...

With a more heterogeneous bunch in charge of creating the internet, we might also see a more multilingual online landscape... The safety tools on the internet might look different too. For example, social media users might be able to verify their accounts without giving up their phone numbers, as Facebook – the most popular platform in the world today – and WhatsApp, which is also owned by Facebook's parent company Meta, currently demand. This could make users less vulnerable to unsolicited messages or having their home addresses and other information that's linked to phone numbers being leaked in hacking incidents.

They might also have the option of using a pseudonym on their accounts... "If you're a white man in Silicon Valley or Silicon Roundabout in London, if you've never experienced anything from small microaggressions up to very severe violence throughout your life, then it's not your natural tendency to think about those things when designing technologies," she says. But women and minorities bear the brunt of online abuse.

Had they been in charge of creating the internet, they may well have prioritised safety measures. And they might have done so from the start. Social media platforms such as Facebook, Twitter and Reddit, for example, now ban revenge porn on their sites. But they only did so in 2015 — roughly a decade since their respective launches — after facing pressure from leading female activists, says Chander. "That should have been the policy from the very start."

None of the platforms the BBC contacted were willing to explain why it had taken 10 years to implement the policies.

Now, admittedly this is all speculative. But is is quite clearly (in my view) intended as such, the exploration of a possibility ("might", "may well have", "could"), not set out as a definitive answer to the question, still less a quantitative assessment of how exactly this would be brought about. Even so, it is speculation founded in a very obvious problem, and the basic postulate that we should have more people who experience the problems described in charge does not seem outlandish to me in any way. Is it the whole answer ? Certainly not ! Is it part of the answer ? I think so. I will address some of the objections raised shortly, but this very basic premise - the idea that if you have people who experience the problems first hand in the management team, you stand a better chance of addressing those problems - seems very robust to me indeed.

First, in what I think is the best part of the article :

Carolina Are, a researcher who studies online abuse at Northumbria University in the UK, agrees. Content moderation "replicates what the patriarchy does, which is ruling what women should and shouldn't do with their bodies, and then punishing them if they deviate from the accepted trajectory of that".

Are is an avid pole dancer who regularly posts pictures and workout videos online. Although she doesn't solicit sex or post full nudity, Are has had her TikTok and Instagram accounts deleted or shadow banned – where content is blocked in a way that is not immediately obvious to the user – on numerous occasions. In a recent occasion Instagram got in touch with Are a day later to say her account was restored after a "content moderation error".

"Sometimes women, just by existing in their bodies, not even performing a sex act, are banned," says Are, who stresses that pole dancing is a sport created by strippers, much to the confusion of social media algorithms...  "What we see on the internet reinforces a lot of negative stereotypes in women, who are often hypersexualised," says Suzie Dunn, an assistant professor of law at Dalhousie University in Canada.

Surely it is indeed men who have the stronger tendency to objectify women, to reduce everything to sexuality. Look, I'll freely admit to this myself - it's very difficult to avoid sexualisation of hot people*; to me, it's a pants-dropping act of sexual provocation, to a woman, it's just a dance. Likewise to me it's just George "oh god I'm so dull" Clooney, to certain women, it's... something else. Is this an innate male tendency ? Perhaps partly, but my suspicion is also that culture plays a larger role than one might think - sexual preferences vary strongly between different cultures, after all. This being true, it would seem to strongly imply that to avoid unfairly labelling women on the internet as being inappropriately sexual requires a female perspective, given the prevailing cultural norms.

* I cannot for the life of me understand the existence of a pole dancing school for children in Prague. No amount of reasoned argument is going to make me accept that that is anything but all kinds of wrong. Perhaps that's my cultural/masculine bias, but so be it. It's just.... urrrgh !

While the article devotes the great bulk of its content to speculations of this nature, that we need these different viewpoints to break the unwittingly but damagingly enforced cultural biases, it does also conclude with a note of caution. Again, the point is to examine and explore the issue, not to provide a robust conclusion - it's a BBC Future piece, not an academic paper, and such exploration is valuable.

But when it comes to imagining an alternative internet run by women and minorities, it's important not to idealise the situation, experts caution. Profit-making – a defining aspect of today's internet – would likely still be a key driving force. Whether you're queer or black, you could "still be a data-extractive capitalist", says Waldman. [My emphasis]

And there's no guarantee that women and minorities will act differently to men if they rose to the top. "If our societies were arranged in a matriarchy or something where they had all the power, they might do whatever they want, damn the consequences," says Hicks. "We could end up with just as problematic a situation." 

An alternative internet "would be a lot more inclusive", adds Abigail Curlew, a PhD student at Ottawa's Carleton University who researches trolling on social media. "But I don't know if it would be utopian."

That's because the majority of the internet's problems stem from the real world. "Social media is just a mirror to society," says Lilian Edwards, a professor of law, innovation and society at Newcastle University. To tackle the issues women and minorities face online, we must address the discrimination and violence they face offline. 

As I've said many times before, it's all feedback. You can't only reflect or only dictate; no media can only report on society without having at least some influence on society. So I disagree that it's "just" a mirror - mirrors are invaluable for self-examination and, ultimately, change.

This, then, is my central point, which I think the article expresses commendably well : diversity is invaluable in formulating a fairer, better system, because different perspectives must be considered, and non-specialist, indirect knowledge can sometimes be extremely valuable. Beyond this point, which I think an important one, the article doesn't venture much. It does stray from its headline remit of "run by women", but it doesn't attempt to design a better internet, reformulate the business model, or solve all the problems plaguing the world. That's too big an ask.

Finally, there are two main objections which I shall respond to only briefly. Firstly, that it's the commercial business model which dominates the result, not the people in charge. Here I agree. However, it's an open question as to whether you could/would have got a better model without diversity, but the basic point that it's the profit motive which is the bigger factor in developing a shitty internet is one I do agree with. Toxic masculinity is a problem, but it certainly isn't the problem. Again though, this is not how I read the article, not the main take-away point I get from it. 

Second, that there are plenty of high-level "diversity" CEOs in the tech sector already. That's as maybe, but replacing a few individual psychopathic profit-driven monsters with a few other, different* psychopathic profit-driven monsters will change exactly nothing (having the internet run by the proverbial Crazy Cat Lady is just not going to work). But that is not what's being asked, because that's a boring question. 

* In my experience every single demographic bar none contains some fraction of truly awful people.

Clearly, and I think explicitly, the article is concerned with whether if more people who had experienced the problems described therein had been in charge, would those specific problems have stood a greater chance of being dealt with more successfully ? Having a few token diversity hires, even (perhaps especially) at the top levels is not going to do anything. Changing a sufficient number, however, even much lower down the ladder, can result in a qualitative (not quantitative !) cultural change. The reason for this is described in Damon Centola's Change, but in brief, it's network effects and tipping points. When you only have a few members of a minority, it's easy to get trapped in a self-reinforcing stereotype : everyone assumes the minority hires are representative of that demographic, and any who challenge that will face a very tough time of it. Hence the Priti Patels of the world, gaslighting everyone into thinking Britain isn't racist because they themselves are contemptible racist fuckwits.

But I digress. I'm very much looking forward to blogging Change, as I steadily work my way through a long backlog of write-ups to do, but I highly recommend it for pretty much anyone. For now, that's my main point from the piece : diversity is necessary, but not sufficient, non-specialist knowledge matters, and given prevailing cultural norms part of that diversity must be gender related. Could you avoid the biases by simply asking the demographics ? Possibly. But it seems an awful lot simpler to actually have them on the team.