"A society that does not represent us": Why AI needs more women

This article was originally published in English

Six months ago, about 15,000 kilometres from Melbourne, Australia, A politician had her breasts enlarged and her top cut short with a generative artificial intelligence (AI) tool built into Adobe Photoshop.

“Note the enlarged breasts and the outfit that became more revealing. I can’t imagine this happening to a member of parliament.“, wrote Georgie Purcell on the X platform.

In Europe, there is also concern about the new technological trend that is making headlines and disrupting the workplace: generative artificial intelligence. These are the computer systems ChatGPT and MidJourney that learn on the fly and perform human-like functions. In May, the European Commission expressed concern about the “increasing spread of sexual content” fakes and ‘deep fakes’ “women-led AI” (a type of generative AI) and listed a series of legislative measures that institutions have enacted to protect citizens.

But Silja Baller, an economist at the World Economic Forum (WEF), believes that if more women became AI engineers, developing the best and most equitable uses and design functions of the technology in the lab, Women would not be as affected by AI“.

“There must be equal opportunities to contribute “We must all have an equal say in shaping our future,” Baller, the agency’s head of diversity, equity and inclusion, told Euronews. “It’s important that all kinds of people contribute to these transformations,” Baller said.

Women are underrepresented in science and technology

The WEF’s most recent report on the gender pay gap found that women’s representation in AI engineering had doubled since 2016rising from 0.09% to 0.20%. However, according to the report, there remains a “significant underrepresentation” in science, technology, engineering and mathematics (STEM) fields, such as AI.

The generative AI revolution is expected to make a lot of money. A Bloomberg Intelligence report estimates that the market will be worth more than a billion euros in the next 10 years. Numerous studies show that it will also have a disproportionate impact on women. This ranges from technology displaying sexist biases to potentially putting women out of work.

Another WEF report, focusing on the top ten emerging technologies, positions AI for scientific discovery to the head“Advances in deep learning, generative AI and foundational models are revolutionizing the process of scientific discovery,” the report states.

Baller believes that women should be educated and transitioning to jobs that are “growing”… but the numbers don’t add up. “If you look at the transition from STEM qualifications to entering the workforce, you see a decline,” he says. “There is already a gap between acquiring a degree and entering the workforce. Then you have further declines as you move into management positions,” Baller added.

According to the WEF gender parity report, women They only represent one tenth of executive leaders STEM jobs, which is a double disadvantage for technological and labor transitions. This is because, as the document states, women “occupy the jobs of lower growth and lower pay which will likely be negatively affected in the short term.”

AI advocate and business owner Caroline Lair told Euronews that the fact that fewer women are in top positions is because, overall, There are very few women working in the field of AI“The lower overall number of women entering and staying in STEM fields creates a smaller pool of potential female leaders,” she said, adding that this “pipeline problem” exacerbates underrepresentation in senior leadership.

Lair runs Women in AI, a Paris-based nonprofit that aims to attracting more women into data science and machine learning. The NGO has 16,000 members in over 100 countries and offers events, mentoring opportunities and support to aspiring female AI engineers. Here’s how the NGO aims to unblock the pipeline.

The business world has a role to play

But the responsibility of offering these opportunities should not fall directly on associations and governmentsIt is up to companies to do the heavy lifting. “Companies need to foster environments that are conducive to women and mothers, offer mentoring and support women’s career development,” she explains.

Otherwise, they risk being “unwelcoming”, “hostile” and “male-dominated”. According to Lair, getting more women into AI jobs starts at school. According to Eurostat, only 32% of all STEM graduates identify as women, with the highest proportion in Romania, Poland, Greece and ItalyCountries with high dropout rates for girls in STEM need targeted support, she said.

WEF’s Baller echoed this view. But she said that even before women enter the workforce, there is a need to build bridges so that after college, be hired for entry-level positions. The European Commission adopted this mentality when it funded the programme Girls Go Circularwhich aims to equip 50,000 students with digital and leadership skills. Generative AI is such a hot topic that the plan’s 2024 conference revolves around it.

For Lair, the risk is too great without the presence of women in the sector, but AI could become an ally.

“If women are absent, that is, if they are not involved in the collection of data, in its verification and in the identification of hidden social biases in it, or in the creation and development of algorithms in which the biases of the creators may also be hidden, this It may potentially reflect a society that does not represent us.“, he claimed.

“However, if women take over this technology to make their voice heard“Like voting rights and equal pay, this is a huge and truly unique opportunity,” he added.