I first met Julie Choo, a techpreneur and author, at the Deep Learning Summit in London. When I learnt of her interest and research in autonomous vehicles I invited her to participate at the Machine Intelligence in Autonomous Vehicles Summit in Amsterdam as well as interview her for our Women in Tech blog series. We interview leading women in STEM to highlight their talents and expertise as well as learn more about how we can all work to make science and technology industries more inclusive
Julie, tell us a little more about yourself:
I'm the Co-founder of Zoomie, the first personal development app that blends behavioural sciences with machine intelligence, and Founder and CEO of Stratability. I am also author of THE STRATEGY JOURNEY® book coming out later in 2017 which includes my research on data as a currency and autonomous vehicles. My career has spanned many of the different disciplines of STEM, working in Tech, FinTech, banking and business strategy.
I guess you can call me a serial founder and inventor. I have always loved tinkering to make new things and start new projects, especially with gadgets and technology. When I was 9, the video head on our overused VCR wore out (after too many reruns of Star Wars!) and I took to taking it apart and fixing it myself with a screwdriver, asking my parents for parts.
This led me to study engineering at university where I did my thesis in VoIP technologies, which was the technology behind Skype’s success. Since then, I’ve developed my career moving from software engineering into banking, FinTech and business strategy.
Until recently, my career followed the corporate route and I’ve worked with C-level executives and directors of banks and large Fortune 500 companies. I spent several years moving from analyst, to architect, to strategist and managing many large teams.
I founded my startup Stratability at the end of 2015 to explore how to blend behavioural science and data science in the workplace. Having experienced and seen many problems myself during my career, I found I was being asked for help and solutions that didn't yet exist. In the past year, while still working on my book and developing training courses on business architecture and operating model design, I started to work with my co founders on Zoomie. Following a year of research, we are currently developing Zoomie, along with a marketplace platform for users to meet career-related service providers, like coaches and mentors, and service seekers, such as recruiters and employers. We hope to release our functional MVP later this year.
Which industries do you envisage will be disrupted the most by the acceleration of machine intelligence?
Careers and Education: Rise of AI is not just about what jobs can be automated or done by robots. We are already seeing AI used as a personal assistant, in the legal professions and beyond. This will only accelerate. There are already many EdTech companies using data to accelerate learning and I see this as a key future trend. Modern education hasn’t caught up with the digital age. We are still teaching children to learn facts and figures. In the future, learning will use data to tailor education more to learners and show them how to learn with a growth mind-set rather than remember by rote.
Behavioural Psychology or Science and Personal Development: I think we will see disruption to coaching and mentoring through AI. Amazon is not just developing Alexa, a voice recognition hub for the home, but putting together a fund for developers working in voice technology innovation. They are beginning to study behaviours, and how AI can respond and interact with our behaviours. AutoEmotive, a research project from the Affective Computing group at MIT's media lab, is already working on detecting people’s emotions in cars to prevent an accident. What we are working on at Zoomie will bring together a wealth of data in this field which I hope is the start of more advancements here as well.
Transport: Autonomous vehicles are a key trend. But they are not something that exists in isolation. As we move more towards autonomous cars we will be forced to look more at the energy needed to fuel these vehicles. This is not just a discussion about renewable or greener energy. KERS technology, the ability to use a vehicle’s kinetic energy from braking to then accelerate, is being used in F1 right now. F1 are also developing new technology for batteries that charge faster. This will disrupt how we build cars as well as how they are driven. Urban planning will change to adapt to more advanced driving styles as the autonomous vehicle industry starts to mature. AI in transport can be a new future, disrupting how we live together in cities and towns, as well as how we travel.
What do you think are the greatest challenges faced by companies integrating machine learning/deep learning methods?
Data: There needs to be large data sets for AI agents to model from. We will see more companies giving away their software/knowledge for free in order to have more people use it and then gain larger and larger datasets. Comma.ai, a startup in autonomous vehicles, is giving away their code and hardware plans for free on Github already to get as wide a dataset as possible for their future growth. We are seeing the rise of data as a currency, a commodity. Sanjeet Choudary tracks this as the move from a Pipeline business model to a Platform business model. Platforms bring together producers and consumers and their value lies, not just in the transactions that take place on the platform, but in the larger and larger datasets that this produces. Facebook is the best example of this, the data that they hold on people across the world is the value of their company.<.p>
Trust: A major challenge is trust. How can you trust the data that is out there? And what about when the AI agent that is running a service does something that you don’t agree with? I recently had an incident where a company on twitter that I don’t endorse, or want to endorse, found and retweeted a deleted tweet of mine from around a year ago. Chatbots and tweetbots are using AI and data to find content for others to retweet, but they don’t have the human nuance to decide whether the retweet is appropriate.
What key factors that have enabled recent advancements in autonomous vehicles?
Energy: As I mentioned before, advances in energy and batteries have begun to make both electric and autonomous vehicles seem like more of an option. Tesla’s network of charging stations might be located at some not-so-glamourous service stations but the fact that there is a network of places you can charge your Tesla car to make longer journeys means people are seeing the cars as a viable alternative, not just a new tech toy. Whilst I think the F1 industry doesn’t really want to help out the driverless car industry, which could see it become obsolete (I certainly hope not being a big fan), advances in tech for batteries means there are some real leaps forward about to happen.
Race for data: With more data both for and from autonomous vehicles, we are seeing faster iterations on new ideas in autonomous cars. MIT AgeLab is offering cash incentives to Tesla drivers to give them access their Tesla driving experience. They are looking to see how new technologies in their cars are used by drivers, what is useful to them and why. This kind of user experience data is key to making next stage developments.
What developments can we expect to see in autonomous vehicles in the next 5 years?
As I mentioned before, autonomous cars will give rise to a new autonomous network, changes to our energy use and needs, and we will need a rethink of urban planning. I think we will see the start of this over the next 5 years as we begin to become used to driverless cars, and see more pilot schemes started in cities across the world.
Governments are already putting in place regulations for hybrid cars. This will lead to more regulation for autonomous vehicles too, as they begin to emerge. Urban and city planning will need to take these into account, just as they have done with electric cars. With more charging points and incentives given to electric car users, so we will see a rise in physical space made for autonomous cars. Pick up and drop off points might be needed, charging points, overnight parking, these are all things we will need to address in the near future. I think we will also see richer networks with data about travelling and transportation. Google maps already has a massive amount of data and we will see shifts in how that is used with driverless cars. Uber and Volvo are partnering to create autonomous vehicles. Partnerships like this make sense and we will see more of them. Uber is building data which can inform the driving of cars but they don’t own taxis or build cars themselves. Whereas Volvo have the experience of manufacturing that is needed to get these cars onto the road.
What do you think are the main challenges facing the autonomous vehicle industry?
We still have a long way to go to make the leap to people in the mainstream or mass market wanting to ‘drive’ or be driven by a driverless car. The experience of driving is a pleasure for many people. Can autonomous cars really become mainstream if people don’t want to use them? What will happen to the petrol heads, the Top Gear fans? Will driving become something that people do as a hobby or purely for pleasure? Will F1 drivers become gladiators driving for our entertainment?
There is such a shift needed in thinking to take driverless cars mainstream. Will everyone own a car? Or will it become a network of cars at our beck and call? As we won’t need taxi drivers, maybe we won’t need bus drivers, there is a lot of tension around ‘replacing’ jobs traditionally done by people with AI. And there is a whole industry around the selling, insuring, maintenance and customisation of cars. Will we still need mechanics, if the car can self-detect its potential problems and simply return to a base station on its own? That hasn’t become a natural transition for us yet. We can’t predict people, we don’t know how they will react to these changes and their reactions can speed up or slow down the rate of technology in these areas. Right now we are seeing early adopters developing this technology and we don’t really know what will happen when the rate of adoption begins to speed up.
I think infrastructure is a real challenge. How long does the battery in an electric car really last? What is in place for charging stations? We are seeing the beginnings of solutions for these but without commitment to infrastructure there won’t be a mass adoption of this technology.
F1 tech is filtering down into mainstream car design but it isn’t happening quickly. As the pace of change starts to accelerate, will they want to continue to share technology that is disrupting the car industry so much? It will certainly be interesting to see if, when and how, autonomous vehicles will cross the chasm from early adopters to the mass market.
How can we inspire and encourage more women and girls to become involved in STEM fields?
STEM shouldn’t be scary, it is fun. I won a STEM based Co-op scholarship to study Electrical Engineering and Telecommunications at the University of New South Wales in Australia by being able to thread a sewing machine in a super fast time. The engineers who interviewed me thought that this skill, together with my interest in designing clothes in high school, was naturally transferable to designing and building networks. My skills in these areas in my early years as well as my general interest in technology were just as applicable and hence transferable to UX and UI design, technology architecture as well as microelectronics.
I don’t think people really recognise what a broad range of industries STEM actually covers. I think we need to look at how we present science to kids. Why do we seem to give science kits to boys and not girls? They aren’t gender specific. They are toys for them to learn, discover, interact with the world around them. From what I see in education I think there is a fear of failure. If we encourage experiments and fun, failing and learning from it, teach more of a growth mind-set early on then we are opening STEM to a far wider range of talent.
I think it is about showing girls that these pathways are open to them. Williams F1 run a day of karting for girls with Rachel Brooks from Sky Sports, to show girls they can drive and have fun too. The girls come away with the idea that they can do it, be involved in that world, and that is a powerful message.
I loved tinkering with things when I was younger, I loved making things, I’m an inventor at heart. I was making clothes, handbags, circuit boards and tech. That tinkering and trying gave me the mind-set to study STEM. Tech is so common, girls should be encouraged to find out how things work and build them themselves too.
Children are sponges, which can be for great for STEM subjects, getting them when they’re young! STEM is also such a broad category, it covers Science, Technology, Engineering and Maths. Accountants, traders and actuaries use maths. Technology is not just about coding, UX/UI is a creative aspect to the ‘pure’ tech of coding. Psychology, design thinking, behavioural science, these are all science subjects but not necessarily based in data. There are so many pathways into STEM, and I think we need to show girls the range of opportunities STEM has to offer.
What advice would you give to someone starting a career in machine intelligence today?
Don’t be fixated with the ‘sexiness’ of AI – it all comes down to data. Data can tell you so much, whatever you are interested in. Using the data route you will be surprised what you can discover. It is important to realise that data is a commodity, which gives it so many more uses and technologies and we can become so much more creative with it. Learn how data and the basics of data architecture work and then move on from there. You can find a niche from there. Machine intelligence can be applied to anything, everything is data.
Are you working in emerging areas of science and technology, or know of someone who is? Suggest women in STEM fields to speak at a RE•WORK event here.
The Machine Intelligence in Autonomous Vehicles Summit, is taking place alongside the Machine Intelligence Summit, in Amsterdam on 28-29 June. Discounted passes are currently available.
Meet with and learn from leading experts in autonomous vehicles, IoT, the connected car, machine learning methods and predictive analytics. Confirmed speakers include Manager Software Engineering, TomTom; Senior Data Scientist, Pirelli; and Associate Professor in Engineering Science, University of Oxford.
Take advantage of our 25% SPRING discount by applying the special offer to the already discounted early Bird pass here.