Sun. Jul 14th, 2024
    women's day

    Hey everyone

    Happy belated International Women’s Day!

    Have you been watching the US Democratic Party Presidential debates? It’s largely been the same rhetoric but one candidate stood out for me and that was Andrew Yang. He has since dropped out but he was running on the ticket of giving everyone a Universal Basic Income of $1 000 every month. His argument was automation is going to take many vulnerable people out of jobs. We can debate the validity of this ticket but that’s not why I wrote about him here. During one of the debates, he was asked if America is ready for a female president. His response was golden. He said, and I quote, “If you get too many men and leave us alone for a while, we kind of become morons.” 😊

    This article is a week late but it is better late than never, right? If you have read my previous ones, one problem I have highlighted before is the potential of bias in A.I. applications due to the data they are trained on. It is easy to think of bias in terms of race or inequality but gender is another huge factor one needs to consider. Because of our history, women were excluded from certain industries and this resulted in datasets that do not represent their views when they should have. This is changing but in some areas at a snail’s pace. So in this article, I want to take time and tell you of five women who are leading in A.I. development. I learn a great deal from these brilliant minds:

    Kay Firth-Butterfield

    No alt text provided for this image

    Kay is the Head of A.I. and Machine Learning at the World Economic Forum. She emphasizes the development of A.I. frameworks that are ethical and fair (makes sense why I put her as my number 1, right?). Most of her work is directed at governments and national bodies but the same principles apply here at Investec. She engages quite openly on LinkedIn and I would urge you to follow or connect with her there.

    Fei-Fei Li

    No alt text provided for this image

    If we are talking big brains in A.I., we cannot do that and leave out Fei-Fei. She is the Chief Scientist of Artificial Intelligence & Machine Learning at Google Cloud, Associate Professor at Stanford and she directs both the Stanford AI and Stanford Vision Labs. If you have ever delved into Computer Vision you definitely know the ImageNet algorithm. Guess who wrote it? As if that was not brilliant enough, she also runs a non-profit that aims to teach AI to under-represented groups because she believes, “Technology could benefit or hurt people, so the usage of tech is the responsibility of humanity as a whole, not just the discoverer. I am a person before I’m an AI technologist.” With over 150 papers published, she is worth following on Twitter where is quite active.

    Carol Reiley

    No alt text provided for this image

    It is easy to get intimidated when other people seem to know more than you do. That was the case for Carol because she only started coding in her first year at university. With determination, she attained her PhD in Computer Science and Robotics at John Hopkins. She is currently the Co-Founder and President of Drive.AI, a company that uses deep learning for self driving cars. You probably know Andrew Ng, one of the gods in A.I. He is the other co-founder of Drive.AI and credits Carol for taking the startup to where it is right now (Andrew is a co-founder for Coursera as well).

    Hua Wu

    No alt text provided for this image

    One of the hottest fields right now is Natural Language Processing (NLP). How do we teach computers to understand language the same way us humans do? This is an incredibly difficult thing to do and we have to thank women like Hua Wu who are making progress in this field. She is the Technical Chief of the NLP Team at Baidu. Baidu is Chinese version of Google so you can get an idea of the task at hand. Mandarin is not widely spoken across the world. I started learning it in university when I was at Wits because I think it is such a beautiful language so I can comprehend how complex it is to translate it to English. For example, because of the current Corona virus outbreak, in English you can say, “Hello Doctor Li. You are not happy today.” but in Mandarin, it is “nin hao li yisheng. ni jintian bu gaoxing” which directly translates to, “You good Li Doctor. You today not happy”. Apart from the numerous characters and pronunciations, the word order rules are different and change depending on the context. Quite complex but beautiful nonetheless and how do you teach a computer to understand that beauty? Well, Hua and team’s algorithms are processing over 100 million translations a day! They are solving a complicated problem at scale. Their dialogues systems, neural machine translation and NLP algorithms have to handle billions of daily searches from hundreds of millions of users. She also built Duer, Baidu’s conversational AI which is powering home assistants and smart IoT devices, akin to Google Assistant.

    Angelica Lim

    No alt text provided for this image

    If you have an interest in Private Equity/Capital, you know about SoftBank, their humongous US$100billiion+ Vision Fund and how they have been splurging money into startups at a rate and in amounts that have caused anxiety in the top venture capital world. Along with understandable investments in Uber and WeWork (unfortunately), they don’t shy away from investing in forward-looking, moonshot edge businesses like Guardant Health, a company that is using Artificial Intelligence to conquer cancer. Internally, they have a robotics arm that is working on robots that can “feel” emotions. This team is managed by Angelica Lim who is combining Computer Science, Neuroscience and Cultural Development Psychology to model robots that can learn like how toddlers learn to develop emotions and physical expressions. Esoteric aspects asides, did you know the most difficult thing for a robot to play chess is not figuring out the next best move but to actually hold the chess piece and move it! This team, led by Angelica is solving some of these physical challenges along with the softer sides as well.

    Rama Akkiraju

    No alt text provided for this image

    The way I look at A.I. and Machine Learning is they are tools to achieve a certain task. Just like the internet, there are times when you do not need to know how to code the algorithms yourself. Do you know how to write TCP/IP protocols? I’m almost sure knowing how they are coded is not necessary knowledge prior to you sending an email via a POP3 server. This is an increasing trend where A.I. and Machine Learning are becoming democratized tools, giving anyone the ability to spin up experiments and use the algorithms to uncover hidden insights in data. One use case of this is in Marketing where employees are using IBM Watson to improve the way we communicate with our clients. It is not necessary for the marketing teams to know the intricate details of deep learning algorithms but they are uncovering valuable insights they wouldn’t have been able to if it was not for women like Rama Akkiraju who is the leader for People Insights at IBM Watson. Her algorithms infer personalities, emotions, tone, attitudes, and intentions from data using linguistic and machine learning techniques. She brings together different disciplines including AI, psychology, sociology, decision theory and consumer behavior. She has done amazingly well and is one of the only few people to be awarded the title of “Distinguished Engineer and Master Inventor” at IBM.

    There is still a long way to go and I have unfortunately left out many more women who are doing amazing work in A.I. Women like Rachel Thomas, the co-founder of Fast.AI. Sadly most of them go unnoticed even though they are changing the world in incredible ways at all levels. Have you watched the movie Hidden Figures?

    It’s a brave new world!

    57 / 100