Is AI At The Cusp Of Making The World A Better Place?

I have been a big fan of Siddha since the time I heard her on Sam Carrington’s TwiML AI podcast. Siddha Ganju is a self-driving cars architect at NVIDIA, a Forbes 30 under 30 enlister and Advisor at the NASA AI Committee.

She is now co-authoring a book under O'Reilly publications on Implementing Practical Deep Learning on Cloud and Mobile Devices. Siddha gives our community members insight into the future of computer vision, advice to women for forging careers in the field of machine learning and AI, her journey, what inspires her and much more!

Namita: Can you tell us a bit about how you got interested in this field and your journey so far?

Siddha: I’ve been extremely fortunate to have amazing colleagues, mentors and professors who have encouraged me a lot to pursue my interests. Throughout my journey, Grace Hopper, Women in Machine Learning, Women in Computer Vision and other such groups have been pivotal to my success. I got started in ML in college in 2012 and from there it has only grown forwards through all the hackathons and conferences I attended.

Namita: You are currently working for NVIDIA on self-driving cars using computer vision. Can you tell us more about your work with computer vision and your opinion on its future applications?

Siddha: Computer vision has had a huge impact from the revival of deep learning and one of the effects is large scale research in self-driving. Self-driving is a 35-year-old concept, originating from Carnegie Mellon University where scientists in the Navigation Laboratory in 1984 used a very simple neural-network and a combination of other inputs to drive at 20mph.

Today, self-driving research consumes petabytes of raw data, has dedicated data centres for data processing and training and has several stages of testing and validation to ensure safety.

The CV community is evermore engaged leading to amazing results for different tasks.

I can’t wait to see what’s next with GANs! When you see the current state of the art, GANs have achieved almost photorealistic results and it’s just amazing. I think everyone in the field is excited and looking forward to working on NAS.

Model explainability and model cards are also up and coming. AI for accessibility is also something that has been gaining traction and a lot of these systems are in real-world deployment and that’s what makes them amazing.

Today, the good thing is that there are so many ways in which people can get started with AI, be it deeplearning.ai or fast.ai or Siraj Raval’s School of AI and many more - all these are giving opportunities to people who are interested. As Neil deGrasse Tyson rightly puts it “The more science belongs to all of us the less likely we are to misuse it” I wonder about the day when we will thank AI for taking our workaholic-ism away and let us get on with the job of being human. That would be cool!

Also Checkout - Demystifying Data Science With Ulrika Jagare (Director of Technology and Architecture with a focus on Data and AI at Ericsson)

Namita: Advice for women on how they can get started and make better progress forging careers in the fields of tech, data and AI?

Siddha: Show up! Throughout my career, showing up has proved to be half the effort, the other half is always working smartly to do the job. Don’t be scared to ask questions, ask for help, and most importantly, don’t feel that you are under the imposter syndrome.

Today there are many modes of learnings, such as fast.ai, deeplearning.ai, Coursera and so many more, which can open multiple avenues for communication and networking. On a similar note, I’m also co-authoring a book under O’Reilly publications that talk about Implementing Practical Deep Learning on Cloud and Mobile devices.

Namita: What are some of the biggest lessons you learned from working with diverse teams that might be useful for other women who aspire to be in leadership positions?

Siddha: Make sure everyone’s opinion is heard and welcome. And, more importantly, make sure you are making your voice heard.

Namita: For those children who have an aptitude for technology, what are your thoughts on incorporating data science, ML and AI in our current education system?

Siddha: Directly or indirectly, school students are already exposed to AI through their phones, social media, newspaper, and now, history.

I remember reading a Microsoft blog that talked about using OCR + AI to decipher JFK assignation files. They took all the files, photos, handwritten notes and absolutely everything else that the US government had made public. It was around several hundred files, which in human time would take somewhere around weeks to read through and understand it. The Microsoft team took all these documents and fed it into their system and soon enough they could see relationships being derived, making sense of all the people involved in the JFK assassination.

Imagine using such a system to understand the works of Shakespeare like Julius Caesar that we used to read in school. On a similar note, the Maltese government has made a robot citizenship test and Carnegie Mellon has a robot census.

Technology is constantly changing, and we must ensure that we remain upbeat. Several schools in my neighbourhood have already started incorporating this. As an example, in my nearby school, Winters Middle School, Davis regularly invites people working in different fields for a weekly seminar where they talk about their work and how it impacts the world.

I went in once to talk about computer vision and artificial intelligence and I was blown away by the awareness and aptitude of the students! We sure have come a long way since my good old school days.

Another example, AI4ALL is an AI summer camp for middle schoolers that teaches students about AI and helps them develop simple projects of their own.

You might want to check out What Dr. Sandhya Kuruganti (Independent Analytics Consultant & Co-Author Business Analytics: Applications to Consumer Marketing.) thinks about data science, ML and AI in our current education system!

Namita: What inspires you every day?

Siddha: I recall reading Andrew Ng’s recent blog post that the coming generation will be the first generation of AI natives and that they will be taking for granted voice-controlled devices, auto-graded homework, customized news feeds, facial recognition, and unprecedented levels of AI-powered automation.

I recall when I saw, Star Wars or the Matrix or such sci-fi movies, I was so influenced, so in love with the idea that this technology was possible, obviously to my naïve brain I learnt later that it was animation, but the imagination just seemed limitless but for the current generation it’s all real. And on the same lines, we should realize that the choices we make today will shape the world of the coming generations.

We in the AI community hold incredible power and responsibility to step up and make the future world better than the one we know and live in today. The opportunity that AI presents is going to make a huge impact, and the fact that we are just at the cusp of this journey is just humbling.