Women in AI and ML
Yesterday, I went to a forum about Artificial Intelligence (AI) and Deep Learning where I witnessed, for the first time ever, a discussion on data science, AI, intelligent computing, programming, deep learning, and industry breakthroughs lead by women in a room of at least 90% women.
I thought to myself, never before have I felt so comfortable in a room discussing technical and complicated brand-new science and technology.
It was liberating.
Hosted by Women in Big Data and NVIDIA, three women in tech fields spoke about what their company/division is up to in regards to AI and Deep Learning and described the various ways we can get involved in these technologies.
Women in Big Data has been around since 2015. It grew from the issues tech companies in the Bay Area were having when trying to recruit women for work. For instance, the American Association of University Women (AAUW) reported in 2015 that:
Only 26% of computing professionals and 12% of working engineers were women in 2013.
Barely 18% of all computing degrees went to women in 2013 (down from 37% percent in 1985).
The percentage of women in big data is much lower.
WiBD began as a small group of women and is now over two hundred thousand people strong. They host Meetups like this one to share knowledge, network, and introduce women into the professional tech world.
NVIDIA is a huge AI and Data Software company. You may have heard of them if you're a big PC gamer and play Titanfall, Overwatch, or Star Wars Battlefront. NVIDIA builds technologies like their new GeForce GTX 1080 Ticomputer powered by their signature Tesla NVIDIA GPU's which are thousands of times more powerful than CPU’s (and cost about that much more, too). These innovations are the kind of technologies you need to support AI development.
AI is something you've likely heard debated in the media and supported by companies like Facebook and Tesla. AI will soon provide us with self-driving cars, advanced healthcare, and even more efficient, healthy agriculture. The general idea is that you can build a program that learns from its experiences over time within a given set of parameters.
For example, a self-driving car knows what a human looks like, but on the road, it comes across a smaller-looking human. It registers it as something new: it learns that what it's seeing is a child and avoids her. Now, the car knows what children look like and can avoid accidents. This process of teaching itself is called Deep Learning.
Renee Yaofrom NVIDIA discussed several industries AI is already helping: banking, healthcare, agriculture, retail, manufacturing, insurance, and recycling. KariBriski, also from NVIDIA, demonstrated how software is used in AI training to prepare for deep learning. Nazanin Zaker, a Data Scientist from SAP Innovation Center Network, spoke about developing software for Machine Learning and tutorials/classes for to the public.
AI in Seven Industries
To give you a better idea of what AI can do, here’s a summary of the seven industries Renee Yao explained:
AI reduces the need for pesticides by running machines that comb over the crop, identify weeds, and pull them.
There exists a device from Athelas that allows you to take a blood test at home to identify blood infections, HIV, and even Leukemia. It is saving countless lives because it eliminates the need to the doctor to check for cancer. This is a cheap way to test yourself at home - we're talking well under $100 - and it is the next step in preventative healthcare.
AI learns how to better protect against intruders. Enhanced security is huge for financial services. AI also learns where you, the consumer, goes. For example, if you go to a car dealership your phone would administer a message asking if you'd like to take a loan out on a car. The loan is pre-approved specifically for you.
Using satellite data, insurance companies can teach an AI to name or label different kinds of roofs. If every house can be labeled according to the condition of the roof, insurance companies can better prepare for hurricanes, floods, and tornadoes.
AI now predicts the next newest thing in fashion using your Instagram profile. Instagram is huge for posts about looks and fashion and new, material products. Using that data (color, size, style, likes per photo, etc.) it predicts what to invent next.
Sorting parts is a tedious job. Sometimes very similar parts are hard to identify. AI aids in the sorting, tagging, and tracking of these objects during shipment to the assembly site.
Very little of our waste is actually recycled 🥺 and a lot of what is recycled by consumers is actually non-recyclable and ends up thrown away, sometimes with other things that usually are recyclable. AI can sort through recycling and trash to fix this issue.