top of page
The  iluli by Mike Lamb logo. Click to return to the homepage
The iluli by Mike Lamb logo. Click to return to the homepage

Does Science Have a Bias Problem?

How objective are you? Most of us like to think that we always make rational and fact-based decisions, but studies show that we are often swayed by unconscious biases.

Thankfully we can rely on science to give us an objective and unbiased take on the world, with its robust, impartial research and cold, hard data. Or can we? It turns out the truth is a bit more complicated…

In a recent iluli video, I explored how science and technology have reinforced inequalities in our society through data gaps – sinkholes that open up when research data is incomplete and not truly representative. These oversights can have huge repercussions. So how do we fix them?

If you’re based in the UK, you are likely to have seen the work of Caroline Criado Perez, even if you might not realise it. In the past decade, she has led two successful campaigns which helped to break new ground – putting the first female face on a banknote (Jane Austen) and the first statue of a woman in London’s Parliament Square (honouring the suffragist leader Millicent Fawcett).

Her 2019 book, Invisible Women: Exposing Data Bias in a World Designed for Men, was a big inspiration for this video. It exposes the hidden biases and inequalities that some of us may previously have been oblivious to and it persuasively articulates why data bias is a problem we should all be concerned about.

It has certainly made me think differently, more of which later…

When default is at fault

Picture a crash test dummy. If you had to attribute some personal characteristics to the dummy, how would you describe it? Does it have an ethnicity, age, disability or gender? You would like to think that the designers of crash test dummies intended for them to be generic stand-ins for any human being, so these types of characteristics would be neutral.

While this approach may be well intentioned, it creates a problem. Human beings come in all different shapes and sizes, so how do we decide on what the generic default should be? As Criado Perez explains, the answer researchers settled on for many decades was ‘Reference Man’ – a Caucasian male, aged 25 to 30, who weighs 70kg. This excludes a lot of the population, and the resulting data gap has real world consequences. When a woman is involved in a car crash she is 47% more likely to be seriously injured.

What’s true of crash test dummies is true across many other aspects of life – from tax systems to office culture to urban planning to disaster relief. Criado Perez argues that the world has been built and designed using data collected on men, ignoring the needs of half the population.

As she succinctly puts it:

“The male experience, the male perspective, has come to be seen as universal, while the female experience — that of half the global population, after all — is seen as, well, niche.”

Is tech making things better or worse?

This failure to account for women and marginalised groups is often the product of institutions producing data that aligns with their own experience.

Take Silicon Valley. The research and development taking place here has a huge impact across the world. You’re reading this newsletter so there’s a good chance that, like me, you’re excited by the possibilities created by emerging technologies which will shape our future.

But how diverse are the teams who pioneer world-changing technologies in this region of northern California?

In Invisible Women, Criado Perez warns that, as a sector, tech has gone backwards:

“If in Silicon Valley meritocracy is a religion, its God is a white male Harvard dropout. And so are most of his disciples: women make up only a quarter of the tech industry’s employees and 11% of its executives. This is despite women earning more than half of all undergraduate degrees in the US, half of all undergraduate degrees in chemistry, and almost half in maths. More than 40% of women leave tech companies after ten years compared to 17% of men.”

Black and Hispanic workers are also significantly under-represented in Silicon Valley firms.

Organisations have more varied perspectives and fewer blind spots when they are made up of diverse mixes of people. This is particularly crucial for the world of tech, where developments are so far-reaching. The very nature of algorithms and AI mean that biases resulting from data gaps are not only reinforced, but amplified.

If this all sounds quite worrying, there are good reasons to be optimistic. When data gaps and inequalities are acknowledged and acted upon, positive change tends to happen pretty quickly. For instance, using more inclusive language in tech job recruitment ads has helped dramatically increase the number of women applying, while significant advances are now being made to reduce bias in AI through the creation of new algorithms which reduce stereotyping.

Applying this thinking to iluli

Top marks to eagle-eyed subscribers who spotted a change in last month’s video – we’ve welcomed a new cast of characters to the iluliverse!

A cartoon image of a diverse range of characters conversing at a party.

It’s a relatively subtle change – the overall iluli style remains the same – but we’ve introduced some more variety in how our iluli people are illustrated. The updated cast members include defining characteristics like hair styles, colour palettes and clearer representation of gender.

Criado Perez’s book got me thinking differently about ‘gender blindness’ and its unintended consequences. Ultimately, the lesson I took from Invisible Women was that we should challenge the instinct to aggregate the diverse mix of humanity into a ‘generic default’.

You may be familiar with a similar ongoing debate around representation of genderand other characteristics in emojis. It’s just one example of how the subject of inclusion is fraught with complexity and why progress can often feel slow, even in areas of life where the stakes are nowhere near as high as they are for designing crash test dummies.

I’m fascinated by the different ways organisations deal with these types of challenges. It’s something I’ll be exploring further this year with a video looking at the concept of ‘just culture’ – watch this space!

Recommended further reading

  • Some Medical Devices Don’t Mean To Be Racist, But They Are (Psyche): Philosophy professors Vanessa Carbonell and Shen-yi Liao make an interesting case for describing objects, like the pulse oximeter which doesn’t work as well on darker skin, as racist. “The things in question are not thoughts, systems, institutions or social practices, though obviously these can be oppressive. Instead, look at the way oppression gets built into mundane material objects.”

  • To Achieve More Equity in the Tech Industry, We Must Reframe Diversity (The Conversation): The lack of gender diversity in the tech industry in the US has got worse. Patricia Baum Salgado, Fellow at the Institute for Social Innovation, writes: “The hostile environment in the tech industry not only has negative impacts on the women themselves, but companies are also missing out. Studies show that more diverse teams are more productive and innovative.”

  • When I Talk to Siri (Flash Readings podcast): This podcast from academics at the Georgia Institute of Technology discusses how the issue with technology’s accent bias is about more than just recognition – it is also about how these devices respond.

Away from data gaps…

Exciting nuclear fusion news!

Back in September we looked at nuclear power and the promise of fusion. Last month, US scientists announced a breakthrough which brings the possibility of a fusion-powered future a big step closer. After six decades of trying, physicists have now found a way to produce more energy from a fusion experiment than was put in. This BBC article includes a good explainer of what it all means.

What I’ve been reading lately

  • Barack Obama’s A Promised Land. Clocking in at over 30 hours as an audiobook, part one of the 44th President’s memoirs only covers up to the mid-point of his first term in office. And yet, the hours in Obama’s company flew by. I was particularly interested in his insights on decision-making and the importance he attached to embracing a diverse mix of views.

  • Tendayi Viki’s Pirates in The Navy: How Innovators Lead Transformation. The title comes from a Steve Jobs quote – faced with the choice of starting a company or joining a large corporation, he said he believed it would be ‘more fun to be a pirate than to join the navy’. Viki argues that you shouldn’t have to choose, and sets out some useful insights on how to drive change and establish innovative systems within large organisations.

  • Matthew Perry’s Friends, Lovers, and the Big Terrible Thing. In my early 20s I went through a phase of coding 24 hours a day while experimenting with polyphasic sleep (I wouldn’t recommend trying it!). I would do this with back-to-back episodes of Friends on in the background, so I’ll always have a soft spot for Chandler and the gang. In this memoir, Matthew Perry tells the alarmingly honest and sad story of what was going on in his life at the time.

  • Chris Hadfield’s An Astronaut’s Guide to Life on Earth. There’s an inspiring story of single-minded determination behind how one of the world’s most famous and charismatic astronauts overcame the odds to get his dream job. I would also recommend his debut novel, an entertaining thriller called The Apollo Murders.


bottom of page