Opinion: An artificial intelligence revolution is under way – but it’s built with gender bias – The Globe and Mail

4 minutes, 4 seconds Read
Open this photo in gallery:

An AI sign at the World Artificial Intelligence Conference (WAIC) in Shanghai, China, on July 6, 2023.Aly Song/Reuters

Laura McGee is the founder of Diversio, a woman-founded AI startup focused on corporate inclusion and diversity. Jérôme Nycz is the executive vice-president of BDC Capital, the Business Development Bank of Canada’s investment arm and Canada’s most active venture capital investor.

Canada stands at the threshold of a revolution in artificial intelligence. Experts agree we are witnessing the beginning of a secular trend on par with the advent of electricity and the internet. However, unlike these previous innovations, AI possesses a unique capacity to shape our culture, reinforce existing biases, and exacerbate inequities.

The algorithms being developed today, and the products emerging from them, are poised to shape Canadian society for decades and even centuries to come. This makes it all the more imperative that the development of AI encompasses a diversity of perspectives that accurately reflects the rich tapestry of Canadian society.

Unfortunately, an examination of PitchBook data reveals a significant oversight: the stark underrepresentation of women in the field. A shocking 90 per cent of all venture capital invested in companies developing AI went to startups with founding teams made up entirely of men, leaving just 10 per cent for startups that include one or more female founders. Companies founded solely by women receive only a fraction of this already small percentage.

This is a significant problem for several reasons. AI is set to revolutionize Canadian society in numerous ways, from medical technology companies using AI for groundbreaking drug discoveries to human-resources algorithms that streamline the recruitment process, and the burgeoning field of virtual reality. We’ve already seen the potential consequences of excluding diversity in these areas: underfunding of women’s health conditions, such as menopause and infertility, the reinforcement of bias in recruitment algorithms, and the sexist portrayal of women in virtual reality. We need more women and other diverse individuals to counteract these biases.

Despite Canada’s prominent position in the AI landscape, we find ourselves channelling support and resources into a surprisingly limited talent pool. This not only poses a significant risk to our competitive edge, but also overlooks a crucial opportunity to enhance our innovation capabilities by broadening our perspective.

Rapley: Yes, artificial intelligence is fuelling a bubble, and it will eventually burst

Internationally, other countries are making strides toward achieving AI equality, with Canada lagging. In 2023, 41 per cent of funding directed at companies developing AI in the United States went to firms with at least one female founder – four times higher than in Canada. In Britain, the figure stood at 20 per cent, doubling that of Canada. The principle of gender equality acts as a flywheel: The more women there are starting and scaling companies, the more female mentors, successful entrepreneurs, and investors we will see, which in turn fosters further innovation and diversity in the industry. Women-founded startups represent an untapped wellspring of potential that Canada cannot afford to overlook.

So, what steps can we take to rectify this imbalance?

First, we must strive to reform AI technology itself, ensuring that training datasets are diverse and representative of all societal groups. For example, there has been a considerable gap in women and girl specific datasets to inform global health programming, exacerbating gender inequality in health care services. The UN Women’s Making Every Woman and Girl Count initiative exemplifies efforts to achieve this goal of closing the gender data gap.

Here in Canada, considerable efforts are made toward a more responsible AI development by AI research institutes such as Mila, Vector Institute and Amii, but we still need to collectively understand and acknowledge the extent of the problem in order to address it.

Second, the investment community must unite in seeking out women-founded and women-led startups, moving beyond the bias-prone pattern-matching decision-making process. It is essential to consciously avoid biases in our evaluation of potential investments, ensuring that male and female founders are asked questions that equally promote their ventures.

Third, the principles of AI should be applied to funding decisions themselves: gathering rigorous data, identifying biases, and implementing corrective measures. In the same manner that BDC Capital started to track and report on diversity, equity, and inclusion portfolio data in 2022, all investors ought to measure diversity within their teams and portfolios, including gender, race, ethnicity, and disability, and aim to make progress over time.

Finally, we must shine a spotlight on Canadian women such as Shelby Austin, Patricia Thaine and Raquel Urtasun, who are pioneering in the AI space, celebrating their innovations and contributions to the field.

It is undeniable that AI will shape the global economy in years to come. What is not guaranteed is Canada’s role in it. We have the opportunity to be a global leader in this space – engaging our entire talent pool will give us our best shot.

This post was originally published on this site

Similar Posts