Everywoman forum 2018: The dangers of non-diverse artificial intelligence

Experts at the 2018 Everywoman forum spoke about the pros of role models and cons of an artificially intelligent future designed solely by men

To prevent artificial intelligence (AI) being biased in the future, the teams in charge of developing them need to be more diverse, experts told the audience at the 2018 Everywoman in Technology Forum.

With the day focused on developing the technology of the future, speakers and panel members explained the importance of encouraging more women and minority groups into the technology industry to ensure current stereotypes in the sector are not inherited by the technology it develops.

Yen-Sze Soon, managing director of Accenture, said this is already beginning to happen as AI developed. Assistants have a tendency to be named after women, whereas those developed to assist in factories are given male names.

Artificially intelligent stereotypes

Soon claimed women need to help to develop AI to put themselves “front and centre of tomorrow’s world”, rather than be forced to use technology that does not fit their needs because they were not part of designing or developing it.

“Women need to take part in this change otherwise social stereotypes will continue,” she said. There is a long history of products not being fit for their audience because they were developed and tested by a small subsection of society.

Tabitha Goldstaub, co-founder of AI community CognitionX, named a few examples including women dying in more car accidents in the 1990s because crash test dummies were only built to male specifications, adverts showing men higher paid jobs than women, and Apple’s Health app originally tracking everything but a person’s menstrual cycle.

“This is only going to be exacerbated by AI,” she said, adding that if women are not involved in building these new technologies situations such as this will continue to be “everyday norms”.

The skills for AI

As AI, robotics and automation begin to make some roles unnecessary, it is clear many of the jobs people will have in the future currently do not exist yet.

Maxine Benson, co-founder of Everywoman, claimed the jobs most likely to avoid automation, such as those requiring adaptability, changeability, communication and collaboration, will provide women with an “opportunity” as these are skills women are more prone to having than men.

“These are all skills that women possess in abundance,” she said.

Many firms are increasingly looking to fill roles with candidates that have both technical and soft skills – “skills that machines will not be able to have” according to Suki Fuller, founder of Miribure.

Some are concerned that in a world full of artificial intelligence, children will grow up with a lack of empathy because of how they are encouraged to interact with it.

This makes ensuring people are still learning to empathise with each other in an artificially intelligent world just as important as developing emotionally intelligent teams to programme and develop human-facing AI.

According to Fuller, because men are “usually lacking in that mega-empathic skill”, women are in a good position to enter the AI space.

“The human aspect, the empathy, the tacit, everything that is the human aspect that quite frankly women seem to have more of, that will be your strength,” she said. “These are the key traits that make us rule the world.”

Fuelling a diverse Stem pipeline

But regardless of whether women are better at developing the soft skills that will be important in the future of AI, the industry still struggles to attract females across the pipeline.

“When you start out in your career or in school, girls are deprogrammed by the time they’re about seven that they’re not supposed to do anything technical,” said Fuller.

Many believe targeting people from a younger age could help to eliminate some of the reasons girls avoid the technology industry, including teacher and parent stereotyping.

Adding computing to the UK’s curriculum to teach people concepts such as computation thinking and coding from a young age was meant to help with encouraging young people, and especially girls, into science, technology, engineering and maths (Stem).

Ensuring how boys are taught and brought up to make them more accepting of working with and being in competition with girls is also important, said Fuller.

Despite these efforts, there are still women who begin their education and careers in Stem but end up leaving the industry further down the pipeline.

Read more about women in tech

  • Tackling the lack of gender diversity in the tech industry could help close the skills gap, but there is a long way to go to solve either, says Eileen Burbidge.
  • The government has committed to sign the diversity initiative Tech Talent Charter in an aim to close the gender diversity gap in the technology industry.

Inma Martinez, venture partner of Deep Science Venture and chief data scientist of Right Brain Future, claimed in organisations the shift needed to attract and retain diverse talent needs to be driven from the top down.

Research has found that having just one woman on a board reduces the likelihood of bankruptcy for a firm by 20%, proving the level of change having diverse representation on a board is capable of.

“Like anything in this world, there are companies that can see the future and move fast, and there are others that are still in 1999,” said Martinez. “It takes one crazy person at the board to say ‘I’m going to do that thing’.”

Many recommend finding a mentor higher up in a firm to gain encouragement and ensure there is a steady pipeline towards these higher-level positions.

“Be awesome and then the mentor wants to mentor you,” said Martinez. “There’s got to be a connection. A mentor normally sees him or herself in you.”

Acting as industry mentors and role models

As well as mentoring, women across the pipeline in the technology industry should make themselves visible to act as role models for those around them. Young women have even claimed they want more encouragement from role models in the technology industry to pursue careers in Stem.

Karen Gill, co-founder of Everywoman, stated being able to see women in senior positions in an organisation is the “single most important change” Everywoman members would like to see in the industry.

“When a woman is able to look up and see what they can be, it can do wonderful things for them,” she said. “Whatever your age or experience, you are a role model for someone ahead of or behind you.”

Many of the experts on the day alluded to the age-old adage that “you can’t be what you can’t see” to demonstrate how important it is to be visible and do your part to encourage change in an industry.

Melissa Di Donato, chief revenue officer of SAP S/4HANA Cloud, claimed she didn’t realise how few and far between women in the industry were until “about six years ago” and that if she had realised sooner she would have done a lot more to try to encourage diversity in the organisations she worked in.

As well as claiming women in the industry have a “duty to each other” to teach, learn and “pay it forward”, she urged women to “take a risk on a young person who has a huge ambition or energy but maybe doesn’t have the tech skills you would traditionally have thought”.

“At every level of your career you are a role model. Even when you don’t realise you’re a role model, you are,” she said.

Read more on Diversity in IT

Data Center
Data Management