LABOUR UNDER ILLUSION OP-ED: The feminine rage against the machine: Why we need women’s scepticism about AI

We had to read it twice and then send it on to make sure we weren’t the only one’s reeling. A major law firm comments on research showing women use AI less than men and warns this could affect workplace outcomes. Their proposed solution? Explain that AI is a useful tool and create psychological safety. And if not, you run the risk of your business’s employment equity goals not being achievable, and your female employees risk becoming even more unemployable. 

Really? If that’s where we’ve landed, we’ve completely missed the point. Our jaws are still on the ground.

As if forced AI adoption in and of itself isn’t contentious (we can write about that next), the very notion that AI literally controls your career is ludicrous. AI ethics then becomes a moot point, even before considering AI facilitating discrimination (be it age, race or gender based). 

The feminine rage against the machine is not a problem to be solved so much as a signal to take notice of – like a sneeze or cough foreshadowing a pandemic. Scepticism in a world led by crazies is rational and valuable.

Scepticism, according to the law firm (and indeed much of mainstream media and overly optimistic politicians) is portrayed as career-limiting ignorance, or stubbornness. As a reason not to hire. As a thing to be overcome by organisations to meet equity targets. Yet scepticism is precisely what the world will lack if everyone were forced into adoption for adoption sake. Scepticism, we argue, is the skill we need more of in future.

Indeed, research shows that women’s scepticism is rooted in the real insights that women are both more risk averse than men (an important balance to reckless, irrational exuberance amid a hype cycle) and in the reality that AI really does pose a greater risk to women’s careers, jobs and tasks (between the types of jobs women are hired into and the invisible labour female employees are expected to take on). Women know that reckless AI adoption is going to cost them jobs – and cost our economy in the long run as the risk roosters come home to rest.

And then there is the other inconvenient truth that reminds us of the social and economic cost of workplace sexism women have to constantly guard against: when women do adopt AI, as good girls doing what they are told to do, they are seen as incompetent by colleagues and supervisors! (Male colleagues however face no such penalty for AI adoption, they are seen as “pragmatic” instead of incapable…)

These are the realities women face, and why scepticism in this case is a sign of the kind of humans you want – need – in your organisation. People who talk disruption without clear strategy and understanding of long-run consequences will cost us all.

We call on leaders and organisations to consider the following before pressuring women to adopt technologies without question:

• Your business relies on having customers. And that relies on your customers being employed. Purchase power is required for any of the new technologies being advocated to actually be bought. Consumers are needed for any product or service organisation, full stop! If AI is doing all the work and we’re training it to do all the work, where exactly are your customers coming from?;

• For every AI licence you’re paying for and forcing your employees to use, you’re also paying for your own company data, employee IP (technically your competitive advantage) and experience to improve the machine. Existing competitors get the benefit of your expertise, while the AI companies themselves are equipped (by your data) to build your replacement. As Satya Nadella, CEO of Microsoft, said recently: “You essentially pay for intelligence twice, once with money and again with something even more valuable: the proprietary knowledge you must reveal to make that intelligence useful. The better you want the model to perform, the more of that knowledge you have to feed it! Models learn from ‘exhaust’, the prompts people write, the tools agents use, and especially the corrections people make when the model is wrong. Every correction is distilled into institutional know-how. This is the kind of knowledge a competitor could never buy and yet enterprises are handing it over”; and

• Finally (for now), let’s call it what it is: surveillance. Forced adoption of tools to create what you hired someone to do with their original, innate capability; creativity; real thinking – now being surveilled to determine how many hours or keystrokes it takes to get work done, or whether they’re in fact outsourcing their thinking to a machine. BS is not baffling the female brain. Worse, employees know all this surveillance is being used as training data to train their replacements – of course any rational worker will, rightly, push back.

Women know this. They know that they’re training their replacements – in fact they’re training your competitors to replace you too. And quite frankly, they’re not that lazy or behind after all.

So what now?

Asking questions is the gateway to the future – even AI relies on you identifying the right problem to be solved and only delivers value when you merge your own ideas about a problem with the right questions to produce a cyborg-type response.

Let’s not focus on the symptoms or the tick-box compliance exercise. Shift to solutionism instead:

• Forced AI adoption isn’t helping you. AI is only useful as augmentation (as authors like Vivienne Ming in Robot-Proof can demonstrate) that relies on the right problems and opportunities being identified (by humans) who leverage the AI with the right questioning to get a more valuable outcome than could’ve been achieved individually. Stop measuring the activity of AI use and start measuring the outcome of value for clients as a result of AI;

• Hire the sceptics, your customers and business will benefit. PS: this must be followed up with you creating an environment they can thrive in;

• Stop hiring managers who aren’t growing and leading people. Managers don’t see hard truths as beneficial, because “difficult questions” are not easy to answer – sceptics, aka high-maintenance employees, are a manager’s nemesis. It’s easier for managers to have little clones doing as they’re told or following the script – but it’s not better for your customers; and

• Can we quickly get onto the AI should NOT be used in recruitment “bandwagon”. We’re creating organisations – whole nations in fact – of surveilled sheep… when we need more sheepdogs!

Why opt for the clones when distinction is management 101 for competitive advantage? DUHHHHH.

And women everywhere: continue to rage against the machine. It’s the right thing to do! DM

Sharon Pearce and Bronwyn Williams are co-authors of Survive the AI Apocalypse: A Guide for Solutionists.