Industry News Details
Interview with Ankit Jain, Data Scientist, Uber - Speaker at Global Artificial Intelligence conference - Jan 2018 Posted on : Dec 23 - 2017
We feature speakers at Global Artificial Intelligence Conference - Jan 17 - 19 2018 - Santa clara - CA to catch up and find out what he or she is working on now and what's coming next. This week we're talking to Ankit Jain, Data Scientist, Uber (Topic : Personalization Using LSTMs)
1. Tell us about yourself and your background.
I am a data scientist at Uber focusing on Forecasting and Self Driving Car’s business problems. Previously, I was leading a data science team for a food delivery startup. I have also worked with Facebook, Bank of America and other companies in a variety of data science roles. I hold a Masters from UC Berkeley and BS from IIT Bombay (India).
2. What have you been working on recently?
I am working on a problem to forecast trips at an individual user level using deep learning algorithms and also working on solving some of the Self Driving Car’s business problems.
3. Tell me about the right tool you used recently to solve customer problem?
Although it varies from problem to problem but I generally tend to use Numpy,Pandas, Scikit learn on top of Python. For deep learning, I primarily use Pytorch/Tensorflow.
4. Where are we now today in terms of the state of artificial intelligence, and where do you think we’ll go over the next five years?
In the next few years, AI will spread it’s wings outside of core technology industry. Currently, we see major AI breakthroughs are achieved by big tech companies like Google,Facebook etc. There are several industries like healthcare, medical records/transcriptions, insurance etc. where there is a lot of untapped potential in terms of data. Although we are seeing some progress currently, but in future it will have massive impact in these industries.
5. There is a negative perception around AI and even some leading technology folks have come out against it or saying that it’s actually potentially harmful to society. Where are you coming down on those discussions? How do you explain this in a way that maybe has a more positive beneficial impact for society?
AI has the potential to transform our lives. We have already seen progress in cases like Self Driving Cars, healthcare diagnosis etc. Any transformative technology when not put to good use can harm the society. For example, when AI is used by hackers to break security and steal some valuable information it’s not a good use of the technology. Ideally, there should be laws and regulations to limit the usage of AI to only those applications which are helpful to the society.
6. When you’re hiring, what types of people are you hiring? The job market for traditional programmers, engineers is very difficult to get into AI space. Are you hiring from that talent pool or is that a different talent pool? In terms of talent, how do you go about ensuring you get the best AI people at your company?
I look for primarily three things when hiring for AI talent. Programming, Stats/Machine Learning and product intuition. Generally, a candidate has to be extremely good in 2 areas and crossing a minimum threshold in 3rd. Also, I tend to focus on diversity to have different perspectives in the team. I wholeheartedly believe in what DJ Patil says “Data Science is a team sport”.
7. Will progress in AI and robotics take away the majority of jobs currently done by humans? Which jobs are most at risk?
All the mundane/repetitive tasks like driving, classifying photos etc. will be taken over by AI. However, all the creative tasks like strategic thinking, creating new music, launching new startup ideas are not under threat by AI (atleast not in short term).
8. What can AI systems do now?
AI is being heavily adopted by all the major tech companies which have access to data, computing resources and great talent. We have seen breakthroughs in image detection, detection of diseases at par with best doctors and even doing things like language translation and driving cars autonomously.
9. When will AI systems become more intelligent than people?
Limitations of AI stems from it’s requirement of big training dataset to make meaningful conclusions particularly in fields like computer vision. Also, AI lacks creativity due to which it can’t create new things that humans are good at. Due to these limitations, it is very hard to achieve what is popularly known as Artificial General Intelligence (AGI). To achieve AGI, there needs to be a fundamental shift in the way we do AI currently.
10. You’ve already hired Y number of people approximately. What would be your pitch to folks out there to join your Organization? Why does your organization matter in the world?
Uber is at forefront of transforming how people and things move around the world. It has provided convenience and employment to millions of riders and drivers across the globe. It is also working on tons of futuristic opportunities like flying cars, self driving cars etc. It’s great time to join a transformative company which will have a huge impact of our lives.
11.What are some of the best takeaways that the attendees can have from your "Personalization Using LSTMs" Talk?
This talk is mainly focused towards understanding of user level behavior based on their past trends and current spend levers like driver incentives. This talk will illustrate innovative ways to calculate elasticity of each user on future spending and how it guides budgeting decisions at Uber.
12. What are the top 5 AI Use cases in enterprises?
Some of the use cases are like Customer Support, Anomaly Detection, Hiring, Credit Scoring and Disease diagnostics in healthcare.
13. Which company do you think is winning the global AI race?
Any company which has figured out ways to collect effective data esp. labelled data will be in best position to win this race. That is why we see major innovations come from big tech firms like Google, Facebook etc.
14. Any closing remarks
Looking forward to a great interactive session with the attendees.