I am tweaking (someone else’s set of) slides for a schools careers talk. Tomorrow Farnborough College have a careers day, and it is my job to tell the students about what physics degrees and careers have to offer. Most of the students will not have a scientist or engineer in the family, as I didn’t when I was their age. So they may have very little idea of what scientists and engineers do, or what careers science and engineering graduates do if they leave science and engineering on graduation, as many do. Tomorrow, I hope to demystify this, and help them make good decisions.
Unlike when I was their age, they have the whole internet with its almost limitless knowledge to call upon. On the whole this is good, they can Google to get the information they need. But no information source is free of biases, and Google is no exception.
To get an idea of what a Googling 17-year old would see, I Googled ‘physicist’ and the then clicked on ‘images’. The result is shown up top. The good news is that the images are not all of white mean, but that is about the only good news. I am scared that 17-year old girls doing science and maths A levels could perform the same Google search and think that physics is not for them.
The first identifiable women physicist is image number 21! Before that there 19 men, and one stock photo image of a pretty young woman doing maths at a blackboard. I despair. In the first 30 images, Albert Einstein alone appears more than three times as often as the entire female half of our species. The image of the one identifiable women in the first 30 images is from an interview with Prof Lisa Randall that contains the line “She is also sometimes referred to as “the physics babe” — a moniker she detests.”. Clearly, I have work to do tomorrow to convince the girls in the audience that physics is for them.
The first 30 images also include 14 images of physicists in front of blackboards. The students in Farnborough College tomorrow are probably going to be disappointed when I don’t arrive trailing clouds of chalk dust.
Google runs on algorithms, on very fancy algorithms that aim to give people the images they want. This is pragmatic and useful, but clearly this can reinforce stereotypes. If many people want a picture of Einstein to represent a physicist then Google will give them images of Einstein front and centre, and so they have the image they are looking for immediately. But if you expect Google to give you representative images of physicists, then this is not what it’s algorithms are designed to do.
It also highlights the fact that algorithms are not regulated in the way that humans are. If someone had curated a set of 30 images of physicist with only one identifiable women, I or someone else would politely tell them that they don’t want to do that. But with algorithms it is harder to see how biases can be corrected. With the ever rising importance of algorithms this is something we as a society may need to think about.
Chance & Necessity 