BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

Do We Need A #MeToo For Machines?

Following
POST WRITTEN BY
Dr Ileana Stigliani
This article is more than 5 years old.

Getty

As if we don’t have enough to worry about these days with concerns that AI will allow machines to rise up against us, now it looks as if they may have become as inherently misogynistic as some of the worst targets of the #MeToo campaign.

This widely held perception doesn’t just seem to be the result of some vague paranoia about technology: we’ve seen numerous examples of how AI can absorb the prejudices ingrained in the information used to train it, and can consequently ‘learn’ to be racist or gender biased. AI has shown itself more likely to associate women with domestic tasks such as cooking, cleaning and washing, while the guys get all the cool stuff straight out of a Gillette advert. And what is potentially really worrying is that AI has a tendency not just to absorb the original prejudices, but to exaggerate them into something altogether more threatening.

So why is this happening? Working with my colleague, Professor Kimberly Elsbach at UC Davis, I set out to find the underlying causes.

Our analysis of the existing research suggests the standard perception of technology is that it’s complex and difficult and, consequently, that it is inherently masculine. But why?

The reason seems to lie in the fact science and engineering have traditionally (and, some would argue, still are) dominated by men. Despite all the advances we have made in the cause of gender equality, it appears we may still be locked into thinking dating back to the 18th century: that machines are somehow ‘masculine’ and consequently need to be made and controlled by men. And this attitude is reinforced by the fact the endorsers of technology are still largely male , which has created something of a ‘boys’ club’ atmosphere across the whole sector.

Getty

The consequences of this male domination are widespread and possibly insidious. For example, a lot of IT hardware and its interfaces are still based on the video game model that arose in the 1980s, when female developers were decidedly thin on the ground. Consequently, many women do not feel as instinctively engaged as they could be.

And even in the new more ‘human’ information technology, it seems our old prejudices are leaking across into the virtual world. Humanoid avatars with reassuringly human voices might make the interface with machines easier and more comfortable, but research suggests that, when dealing with a virtual professional, such as an AI-generated medic, most people would still find a male version more trustworthy than a female one. So much for progress.

So does this all really matter? Should women accept that, given the ongoing shortage of female talent across the STEM arena, the situation is just what it is?

I think the answer to that question is an emphatic ‘no’.

Organisations need to recognize that a failure to invest in redressing the general talent imbalance in the technology sector could have severe consequences. Why? Because, as technology becomes ever more pervasive and integral to our everyday lives, there needs to be more and more emphasis on the deployment of empathy, emotional intelligence and the whole spectrum of soft skills – areas where research shows women can excel. If the partnership between humans and machines that so many informed commentators advocate is to become a reality, then the input of women to its development is not just a nice idea, but an absolute necessity.