Why Women in AI?
The quest for AI began over 70 years ago, with the idea that computers would one day be able to think like us. Ambitious predictions attracted generous funding, but after a few decades there was little to show for it, but, in the last 25 years, new approaches to AI, coupled with advances in technology, mean that we are now realising those pioneers’ dreams.
Today academics and industry leaders are training machines to 'Think Like Us.' They are now able to navigate a map with real-time traffic, spot cancer earlier than a human eye, predict crop yields, trade stock, teach in school, beat a human at Jeopardy and the game Go, schedule meetings, paint an (unnoticeably) fake Van Gogh, predict social unrest, troll people on Twitter and influence elections.
I believe it’s vital that the 'Us' in 'Think Like Us' is representative. We need a call to arms, a movement as powerful as LeanIn and SheCanCode - to continue their essential work and make clear the life-and-death imperative to get women into Artificial Intelligence, and teaching machines to think like us all.
Men currently dominate the digital revolution. The workforces at the biggest tech companies are on average 70% male - women only make up 16% of technical positions and 23% of leadership roles. If this continues we are at risk of our machines being trained to think like intelligent men, rather than intelligent humans.
Traditional software development focuses on building applications to perform well-defined tasks, whereas building artificial intelligence application requires teaching a computer to draw conclusions from provided information and act accordingly. For example, a traditional software application for drawing cats could only copy the images provided to it. However, an artificially intelligent application would identify the features making up a cat and learn to arrange them in the correct way to produce an image of a cat. This means that in simple terms, Artificial intelligence relies on giving a machine the right training data to learn for itself. If we teach a machine on patriarchal or biased training data we are in for trouble. For example Microsoft’s bot Tay was trained on the Twittersphere and it took less than 24 hrs for the AI to become a racist, sexist bigot.
Empathy, nurturing, listening, multi-tasking, intuition, teaching and mothering are skills and qualities we need to be involved in training AI machines. The question is how to ensure they are at the table.
Skills needed to work in AI are typically described in Venn diagrams that show an overlap of Maths, Computer Science, Psychology, and Storytelling. There are many ways for women to get involved, from being a Data Scientist to just ensuring that they are prepared for the everyday world of working with a machine. If you want to find out more, we suggest you attend CognitionX's upcoming event on this very topic -the first of many to come!
Tabitha is the Co-Founder of CognitionX and one of our very own role models for women in tech.
CognitionX is a Community and Directory for All Things AI. Their mission is to bring clarity to the complex world of AI and provide an online tool and physical events to help companies to deploy AI with less risk.
This blog was first published by Tabitha and has been republished on SheCanCode with the author's permission. For that, we thank her.