Industry News Details
Interview with Dan Shiebler, Machine Learning Modeling Engineer, Twitter - Speaker at Global Artificial Intelligence Conference Posted on : Sep 28 - 2017
We feature speakers at Global Artificial Intelligence Conference - Oct - 23 - 24 2017 - NYC 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 Dan Shiebler, Senior Data Scientist, TrueMotion(Topic : Real World Data Science Strategy)
1. Tell us about yourself and your background.
I work as a Machine Learning Modeling Engineer at Twitter Cortex, Twitter’s internal Machine Learning team. In my spare time, I do deep learning research with the Serre Lab at Brown University, where I develop neural networks that can “think” more like humans. Previously, I worked at TrueMotion, where I developed machine learning algorithms that operate on smartphone sensor data to predict a driver’s likelihood of getting into an accident.
2. What have you been working on recently?
At the Serre Lab, I’ve been working on a novel training procedure for convolutional neural networks that tries to match the network’s attention to human attention by manually driving the network’s input layer gradients towards human attention maps.
3. Tell me about the right tool you used recently to solve customer problem?
At TrueMotion, I used neural networks to identify the moments when people interacted with their phones while driving.
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?
Today AI can reliably accomplish single tasks in a vacuum. In 5 years we’ll probably be about at the same place, except the single tasks will be more sophisticated. In 10 years we may have some AI systems that can learn to accomplish multiple tasks.
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?
The main harm of AI to society is its potential to displace jobs. It’s benefit to society is essentially everything else - AI can help us develop life saving pharmaceuticals, assist with medical procedures, give us insights into history, save us time from menial tasks, make dangerous tasks like driving safer…
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?
The best AI people need to be expert programmers who can build systems to move and manipulate huge amounts of data, as well as mathematicians and statisticians who can understand the fundamentals of the algorithms they use and construct the optimal approach for each task.
7. Will progress in AI and robotics take away the majority of jobs currently done by humans? Which jobs are most at risk?
Any job that can be automated will eventually be automated. Some high paying, well respected jobs that will fundamentally change over the next 20 years because of AI include radiologists, advertising executives, and stock traders.
8. What can AI systems do now?
AI systems can play video games, assist with biological simulations, process and derive insight from images, audio recordings, and text, drive cars, interact with humans through language in a relatively coarse manner, process motion sensors, allocate resources, and do a million other things.
9. When will AI systems become more intelligent than people?
Essentially never.
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?
Twitter gives everyone the power to create and share ideas and information instantly.
11. What are some of the best takeaways that the attendees can have from your "Automated Schema Matching and Data Unification with Smart ETL Services" talk?
Attendees to this talk will learn about the differences between developing algorithms in an academic or competition setting versus in an industry setting. They will also learn about the common pitfalls of industry data science and how they can avoid them with a practical, goal oriented approach.
12. What are the top 5 AI Use cases in enterprises?
- Medicine and Healthcare
- Transportation
- Agriculture and Farming
- Science and Humanities research
- Advertising
13. Which company do you think is winning the global AI race?
I think there are a lot of top contenders in different spaces. The tech giants have the strongest presence in academic conferences, but there are a lot of very competitive and secretive companies, like hedge funds, that are also working on cutting edge AI research.
In general, today’s techniques become more effective as you pour more more data into them, so companies that have access to more data will be able to build more powerful algorithms.
14. Any closing remarks
I hope that you enjoy my talk!