Stories In Code

Code Blue: The Disappearing of Women (and Others) in Tech Is a Critical Threat

Since Covid, the number of women and those from ethnic minorities being hired in the tech industry has taken a disturbing downward turn. Even AI sees this as a problem…

We all use software. You’re using it now as you read this. Call it an app, call it a UI, call it a database, call it AI, call it infuriating. Call it whatever you want.

In the back offices of the software industry, a stark silhouette emerges, delineating a landscape marked by a conspicuous absence of gender and ethnic diversity. The virtual streets of this digital metropolis are predominantly marked by male footprints, with women and marginalized groups relegated to the soft-skills basement, their presence sparse and their voices often drowned out in the cacophony of male-dominated discourse.

A report from the National Center for Women & Information Technology (NCWIT) paints a vivid picture of this disparity, revealing that women occupy a mere 27% of professional computing occupations in the U.S. workforce and that of that less than 3% were black or African American (NCWIT, 2023). The glass and concrete jungles of major tech companies are no different, as evidenced by Google’s 2023 diversity report, which discloses that that the hiring of women in the tech industry has dropped from 37.5% in 2022 to 33.8% in 2023 (Google, 2023). This trend is also reflected in the UK where the number of women in technology jobs has dropped by around 20,000 while male participation has risen by over 100,000 between 2022 and 2023 (Bleach, 2023). Women in tech are becoming an endangered species (yet again). The picture is similar for ethnic minorities…

OK, we are told that diversity is a thing, but why should I care about who writes software? Programming languages are just the gobbledygook of the shadowy unsociable figures that make the machines do stuff. Mostly harmless. Totally boring.

Are you sure about that?

In the labyrinths of the digital realm, software stands as more than mere strings of code — it is the embodiment of human cognition, capturing the essence of our thought processes, values, and cultural norms (Winograd & Flores, 1986). Software developers weave together complex code to solve intricate problems, their creations reflecting their understanding and interpretation of the world. But, as with any good story, there are layers of complexity that belie the surface.

Like the smoky rooms and shadowy corners of a noir detective story, software harbours its secrets and biases, reflecting the beliefs and prejudices of its creators and the society in which it was forged (Friedman & Nissenbaum, 1996). This digital realm encapsulates and reflects our society — for better or worse.

I remember a colleague at a work pulling out his phone and proudly showing off an app that could turn your face into a fantasy character. We all got turned into goblins, wizards etc, except for the one black member of our team. The facial recognition didn’t recognise his face. It’s often not good with the faces of women either. This is the same tech, by the way, that is used by police, security organisations and authoritarian states. And it can’t reliably tell one black person apart from another.

But it’s not just a one-way street. The softwarescape has its way of bending us to its will, shaping our thoughts and actions in the process. In many domains it has become our de facto teacher and confidant, guiding us through complex tasks from accounting to making films (everything from TikTok to features) to guiding us through education (all those e-learning platforms) and so much more… The design of its interfaces and the algorithms guide our attention, manipulate our decisions, and even warp our cognitive processes (Norman, 2013). Think PowerPoint and your slide with 3 dot points or your ability to remember anything without your phone (phone numbers, directions, dates etc.)

And let’s not forget the gamified experiences that lurk around every corner, turning our daily interactions into a game of points and rewards, likes and emojis, driving us to act in ways we might not have otherwise considered (Deterding et al., 2011). This behaviour has clearly improved the quality of our relationships as studies are now emerging showing that in the age of social media we are feeling increasingly lonely and isolated (Bonsaksen et al., 2023).

Our world view is unconsciously shaped through the lens of the software we use, so we might want to keep an eye on who is actually being hired to design and make it. Do these people represent you, your experience, and concerns? If not then the ideas and processes written into the software you use will represent their values, their beliefs, their world view.

There’s also a new kid on the block: generative AI and it writes pretty good code. It’s likely to stir things up, but will it replace the programmers? Here’s what ChatGPT 4 had to say about the role programmers and AI play:

While generative AI has shown impressive capabilities in generating code snippets and assisting with certain programming tasks, it is unlikely to completely replace programmers in the foreseeable future. Here are some reasons why:

- Complex Problem Solving: Programming is not just about writing code; it is also about solving complex problems, understanding the requirements, and designing solutions that meet those requirements. AI lacks the creativity and problem-solving skills that human programmers possess.

- Context and Communication: Human programmers are able to understand the context of a problem and communicate with stakeholders to clarify requirements and make informed decisions. AI lacks the ability to understand context and communicate effectively with humans.

- Ethics and Responsibility: As AI systems can be biased and make errors, there is a need for human oversight to ensure ethical considerations are addressed, and to take responsibility for the outcomes of the software.

- Learning and Adaptation: While AI can learn from data, human programmers can learn from experience, adapt to new situations, and apply their knowledge in diverse contexts.

- Creativity and Innovation: Humans are capable of creativity and innovation, which are essential for developing new technologies and solving new problems. AI lacks this creative capacity.

Generative AI can be a valuable tool to assist programmers by automating repetitive tasks, generating code snippets, and providing suggestions. However, the role of the programmer will likely evolve to involve more collaboration with AI systems, rather than being replaced by them. Programmers will still be required to guide the development process, make critical decisions, and ensure that the software meets the necessary requirements and ethical standards.

Thanks for that ChatGPT4, but in the current climate I’m wondering whether you’ve set the bar a little too high for the Hoomans…

Paradoxically, ChatGPT probably represents more diversity than its creators in the tech sector because of the amount of data required to train it. So, let’s play a game, a little illustrative thought experiment. Let’s take ChatGPT’s response and see how the story changes with just a flick of the pen, and some simple word substitutions.

While predominately male human coders (MHCs) have shown impressive capabilities in generating code snippets and assisting with certain programming tasks, MHCs are (hopefully) unlikely to completely replace the diverse population of the world (DPotW) in the foreseeable future. Here are some reasons why:

- Complex Problem Solving: Programming is not just about writing code; it is also about solving complex problems, understanding the requirements, and designing solutions that meet those requirements. MHCs lack the creativity and problem-solving skills that the DPotW possess.

- Context and Communication: The DPotW are able to understand the context of a problem and communicate with stakeholders to clarify requirements and make informed decisions. MHCs lacks the ability to understand context and communicate effectively with the DPotW.

- Ethics and Responsibility: As MHCs can be biased and make errors, there is a need for DPotW oversight to ensure ethical considerations are addressed, and to take responsibility for the outcomes of the software.

- Learning and Adaptation: While MHCs can learn from data, the DPotW can learn from experience, adapt to new situations, and apply their knowledge in diverse contexts.

- Creativity and Innovation: The DPotW are capable of creativity and innovation, which are essential for developing new technologies and solving new problems. MHCs lack this creative capacity.

MHCs can be a valuable resource to assist the DPotW by automating repetitive tasks, generating code snippets, and providing suggestions. However, the role of the DPotW will likely evolve to involve more collaboration with AI systems, rather than being beholden to the error prone outputs of MHCs. The DPotW should be made a requirement to guide the development process, make critical decisions, and ensure that the software meets the necessary requirements and ethical standards.

That’s more like it….

We should not be complacent about who works in the tech sector. The quality of our lives and livelihoods depend on it.

I’ll give ChatGPT4 the final world as I asked it about the impact a lack of diversity might have on the tech sector going forward:

In the fast-paced, ever-evolving world of the tech sector, the lack of gender and ethnic diversity isn’t just an oversight — it’s a significant detriment to the industry’s full potential. Like a garden without a diverse range of flowers, the absence of different perspectives and ideas stifles the growth and bloom of innovation and creativity. It’s not merely a matter of filling quotas or checking boxes; it’s about enriching the tech ecosystem with a multitude of voices and experiences that can drive the industry forward to new heights. Diversity isn’t just a nice-to-have; it’s a must-have for a thriving, dynamic, and innovative tech landscape.

References:

Bleach, T. (2023). Number of Women in UK Tech Falls by Over 20,000; Despite Increase in Male Employees by Over 100,000. The Fintech Times [online]. Available from: https://thefintechtimes.com/number-of-women-in-uk-tech-falls-by-over-20000-despite-increase-in-male-employees-by-over-100000/ [Accessed 26 October 2023].

Bonsaksen, T., Ruffolo, M., Price, D., Leung, J., Thygesen, H., Lamph, G., Kabelenga, I. and Geirdal, A.Ø. (2023). Associations between social media use and loneliness in a cross-national population: do motives for social media use matter? Health Psychology and Behavioral Medicine. Vol. 11 №1. p. 2158089 [online]. https://doi.org/10.1080/21642850.2022.2158089.

Deterding, S., Dixon, D., Khaled, R., & Nacke, L. (2011). From game design elements to gamefulness: defining” gamification.” Proceedings of the 15th international academic MindTrek conference: Envisioning future media environments, 9–15.

Friedman, B., & Nissenbaum, H. (1996). Bias in computer systems. ACM Transactions on Information Systems (TOIS), 14(3), 330–347.

Google. (2023). Google Diversity Annual Report 2023. Retrieved from https://diversity.google/annual-report/#!/home

Norman, D. A. (2013). The design of everyday things: Revised and expanded edition. Basic books.

Winograd, T., & Flores, F. (1986). Understanding computers and cognition: A new foundation for design. Ablex Publishing.

National Center for Women & Information Technology (NCWIT). (2023). Women in Tech: The Facts. Retrieved from https://wpassets.ncwit.org/wp-content/uploads/2021/03/19193926/05_19_2023_BTN_FullSize.pdf