AI & gender: Beware of embedding past prejudices into our future

In conversation with Dr Tomas Chamorro-Premuzic, Chief Innovation Officer, ManpowerGroup, and Professor of Business Psychology at UCL.

At the recent City & Financial WoMen in Finance summit, Vivienne Artz OBE, an advisory board member and past president of Women in Banking and Finance, spoke to Dr Tomas Chamorro-Premuzic, Chief Innovation Officer, ManpowerGroup, and Professor of Business Psychology at UCL, about the likely impact artificial intelligence will have on the workplace, and especially on the gender balance within it.

AI and the workplace

Artz opened up with questions that touched on how we think about achievement, expertise and merit.

Is AI forcing us to re-evaluate how we work, learn and think?

TCP: Most AI is redefining the meaning of expertise. It used to be having the answer to a lot of questions. Now it is more about asking the right questions to begin with and knowing how to treat the answers that AI generates, particularly being able to vet the answers that come from AI and are euphemistically known as “hallucinations”. AI is a big threat to humans as it can make up stuff which historically had only been in the realm of human imagination. Knowing how to act on the basis of those insights also means knowing what to ignore is crucial.

The real disruption is coming from how we humans demonstrate our expertise to others. Most people in the room are in the business of selling their judgment, and if a lot of the answers to questions can be crowd-sourced and commoditized then why should people pay extra for the training and education, experience and expertise that we as experts have?

How is AI changing the workplace for women?

The real opportunity is for AI to increase meritocracy in the workplace and society and that’s why a lot of people resist it. You can think about AI as something that can make workplaces more data-driven and that will also emphasize the need to hire and promote people, especially leaders on the basis of soft skills. If you did that in a gender-blind way, and you actually focus on the unique human skills that make people effective at work, things like: emotional intelligence, empathy, self-awareness, humility, cultivability and integrity, you would actually end up with more women than men in leadership roles and obviously that’s very disruptive.

“If you … focus on the unique human skills that make people effective at work … like emotional intelligence, empathy, self-awareness … you would end up with more women than men in leadership roles … that’s very disruptive.”

Dr Tomas Chamorro-Premuzic, Chief Innovation Officer, ManpowerGroup

Those that are the status quo, mostly men, have little desire to disrupt themselves which is why I think there is this backlash. There is a lot of resistance to AI in most organisations even though everyone would benefit from being more meritocratic and talent centric.

AI and skills

Discussion moved to how AI would be deployed, and the question of whether it threatened to reduce the number of jobs available to human beings.

Do you think AI will increase or decrease the skills at work?

There’s an opportunity to level the playing field. A study published by MIT a couple of weeks ago showed that young, not-so-educated, not-so-skilled novices or amateurs, across 10 different areas of performance using Chat GPT can reduce the performance gap with experts in those fields by 50% (when the experts do not use generative AI or aren’t leveraging AI for performance).

All the fears and sensational headlines that AI will kill all jobs and make humans useless and obsolete may be ill-conceived. We are less likely to lose our jobs to AI than to other humans who are using AI. Complacency and over confidence may lead experts to stagnate, thinking they can rely on their own expertise and intuition, which may in turn improve things for new entrants into the field.

How do we keep AI under control?

We can’t keep AI under control – this is both scary and exciting. Attempts to stop the research makes little sense. We need to progress AI as fast and furiously as we can. Regulation is needed to show that the application is not harmful and that it is practical. No organisation or individual has the answers to where this might lead.

If we accept that the status quo in any area is far from ideal and is generally a low bar then we need to encourage a healthy degree of experimentation. Horror stories of AI going bad generally exposes the bias in organizations for example: when one self-driving car crashes and people are shocked and scandalized but they are perfectly OK with 1.3 million people dying every year worldwide courtesy of human drivers.

“We should have faith in human ingenuity and creativity to actually push the boundaries of human imagination.”

Dr Tomas Chamorro-Premuzic, Chief Innovation Officer, ManpowerGroup

Looking at generations, I have hope in younger people as they are either digital natives or more comfortable with technology. ChatGPT went from 0 – 100 million active users in two months – faster than Tik Tok and Instagram. It shows you that there’s a lot of appetite and demand for actually utilizing / automating tasks. We shouldn’t be spending our time on boring, predictable, repetitive, standardized tasks. Will we reinvest the time we save doing creative things or will we waste it on YouTube and Tik Tok – that’s to be determined – but I think we should have faith in human ingenuity and creativity to actually push the boundaries of human imagination.

It will be disruptive, it will require effort in short term but an exciting opportunity to make work and life in general more interesting.

AI and leadership roles

Much has been said about the dangers of AI replicating human biases but, in a pretty blunt assessment, Chamorro-Premuzic made the point that it can cut both ways.

Why is it so hard for competent people, especially competent women, to advance in the workplace?

Most people are interested in helping competent women to become leaders. Usually when we look at helping women who lack confidence with strategies to make them more assertive, we miss the fundamental issue. If we wish to make competent women become leaders we should start with annexing incompetent men who are in leadership roles.

It makes me laugh when people treat diversity or gender interventions as meritocracy or positive discrimination when there is in fact this positive discrimination today – if you are an over-confident, narcissistic and underqualified man you actually have more chances to get the top. This doesn’t just make it harder for competent women to get to the top but also competent men. Men who display empathy, humility and integrity are actually overlooked as they don’t resemble or meet the traditional archetype of the old, masculine but not very smart incompetent leader.

Make diversity and inclusion more data driven as intentions alone don’t drive progress. Focus on outcomes and results.