The Plough, the Smartphone, and Digital Intent

Thoughts on re-orienting bias in technology

We tend to think that technologies don’t, unlike people, have biases. After all, fire burns the same for everyone. A knife cuts for whoever wields it. And the internet - or an app on it like YouTube - enables video watching for all.

But that’s not true. At all.

Thinkers including Doug Rushkoff discuss how even seemingly ‘democratic’ media can have inherent biases. Intentional or not, bias can be a feature of a media’s design, and not a bug. I not talking about biases in the text, software, data or algorithms that run on a media. I am talking about biases baked in to the medium itself, such as in to a book, a door, or a VR headset.

One standout example is a fascinating piece of research on an old, old media - the agricultural plough. It’s called the ‘Plough Hypotheses.’ And I quote this article …

“ … economists found a major difference between women's roles in societies that were descended from farming communities that used ploughs and those whose ancestors used hoes.”

And …

"The descendants of societies that traditionally practised plough agriculture, today have lower rates of female participation in the workplace, in politics, and in entrepreneurial activities, as well as a greater prevalence of attitudes favouring gender inequality."

The analysis was done from studies of more than 1,200 different language groups from around the world. That’s a lot! Not only did they find differences in social behavior, but also in attitudes. For instance, “the authors found that societies and ethnic groups descended from plough-using peoples were significantly more likely to agree with statements that men should have first choice of jobs and that men make better political leaders. These attitudes persist even when the people have emigrated to western nations.” (link to quote.)

Essentially, the plough hypotheses suggests: the media was biased in favor of men - which led to a division of labor - which led to broad cultural beliefs about the roles of men and women in society.

A futurist gazing down at the plough in 3000 BC would have marvelled at its impact on the 'future of work'.

(Note: The plough hypotheses rests on the assumption that men are the physically stronger gender. This notion is being challenged by recent research, which I would love to explore in another volume.)

Back to that bias conversation.

The Un-Biasing of Media?

For a means of production to carry an inherent bias turns out to be a big deal in the long run. Same for a means of communication, or a means of learning, or of play. What about the advanced technology of today then? Have smartphones evolved in this sense? Are they increasingly ‘un-biased’? The hard answer, depending on whom you ask, is mostly a resounding no! ….and a feeble yes.

The smartphone actually has a long list of biases. Topping the list: you need a thumb, you need to be able to read, and you need to be online.

Is the smartphone unbiased in its function as a container and distributor of information? Jony Ive once said about the iPhone: “Everything defers to the screen.” This emphasis on ‘removing all but the very essential’ drives today’s technology design mindset. The goal is that the blank slate can then be adapted and shaped purely through usage.

The smartphone has this spare, flat, paper-like form. It may well be the last tangible technology that defers to an unbiased blank slate. Even as digital technology gets smaller and more intimate with the human body, inherent biases seem to emerge. A digital watch or ring - fitting the fingers or wrists of some wearers but not others - have biases of form eerily akin to the ancient plough.

Could a mainstream wearable device of the 2020s end up creating advantages for humans of specific gender orientations, ethnic-linguistic minorities, or cognitive capacities?

Media as Intent

A media’s biases shaped through usage is disruptive to human behavior on a wild level. Digital media's immanent biases tend towards the micro-local, discrete, episodic, tailored, and interactive. Bias within digital media morphs into something ephemeral, complex and altogether unrecognisable. Bias in digital becomes the interface to human intent. And in the process becomes almost indistinguishable from intent itself.

Thinkers including Kevin Kelly and the wonderful James Lovelock have proposed that digital technology has an intrinsic intent - and that they may be an extension of life itself. A captivating notion! But as Doug Rushkoff illuminates in his talk below, this notion could distort our ability to distinguish between the essential ‘figure and ground’ - between the technology and the human.

In my work as ethnographer and behavior designer, I relish looking at ‘digital intent’ in everyday, interventionist products and services.

Indian startup Meesho, for example, is a clever app. It creates a new category of e-commerce player called a ‘Reseller’. A reseller earns a commission by curating and selling products to downstream customers. Meesho lets the reseller do this within WhatsApp, Facebook or other social media, through direct tech integrations. The reseller can add their own profit margin to the product’s price. This limitless profit potential means that resellers with deep demand understanding, exeptional curatorial skills, and sales savvy to match can make extraordinary gains.

No wonder Meesho is growing at a crazy clip. Interestingly its user base is 80% female. One reason for this may be that Meesho’s biases advance the unique skillsets and behaviors of women cultured in India. These include cooperative networking, socialising as a way of life, proactive communication, well-honed aesthetic sensibilities, homemaking skills, and many many more. In the wake of income uncertainty via social distancing, Meesho’s online-only bias may actually favour women, who tend to be home-bound way more than men in India.

Meesho thus becomes an accelerator of income equity for women in a very earning-unequal society. Indeed the app feature in the store says it all.

Digital biases should serve positive human intent, even when such intent is only contextual and momentary.

The algorithmic layer in digital introduces new complexities to this happy alliance between bias and intent. Algorithms can enhance - or invert, subvert, or just plain disrupt - the media bias. Troubling signs of algorithmic gender bias are already evident in many areas of mainstream technology. And we are only at the beginning of the beginning of this saga. I hope to dig more into algorithmic bias in future volumes.

Until then: check your privileges, and take care.