[ad_1]
Wing himself never considered studying computer science. In the mid-1970s, inspired by his father, who was a professor in that field, he entered MIT to pursue electrical engineering. When she discovered her interest in computer science, she called him to ask if it was just a passing fad. After all, the field didn’t even have textbooks. He assured her that it wasn’t. The wing changed branches and never looked back.
Formerly corporate vice president of Microsoft Research and now vice president of research at Columbia University, Wing is a leader in promoting data science across multiple disciplines.
Anil Ananthaswamy recently asked Wing about his ambitious agenda of promoting “reliable AI”. 10 research challenges It found itself in an attempt to make AI systems fairer and less biased.
Q: Can you say there is a conversion in the way the calculation is done?
A: Absolutely. Moore’s Law has taken us a long way. We knew we were going to hit the ceiling for Moore’s Law, [so] Parallel computing came to the fore. But phase shift was cloud computing. The original distributed file systems were a kind of baby cloud computing where your files were not local to your machine; they were somewhere else on the server. Cloud computing takes this and amplifies it even more when data is not close to you; The account is not near you.
The next shift is about data. For the longest time, we focused on cycles, making things run faster – processors, CPUs, GPUs and more parallel servers. We ignored the data part. Now we need to focus on the data.
Q: This is the field of data science. How would you define it? What are the challenges of using the data?
A: I have a very short description. Data science is the study of extracting value from data.
You can’t give me a lot of raw data and I press a button and the value comes out. It begins with collecting, processing, storing, managing, analyzing and visualizing data and then interpreting the results. I call it the data lifecycle. Every step in this cycle is a lot of work.
Q: When you use big data, concerns about privacy, security, fairness, and bias often arise. How to solve these problems, especially in artificial intelligence?
A: I have a new research agenda that I’m promoting. I call it reliable AI inspired by the decades of progress we’ve made in reliable computing. By reliability we generally mean security, reliability, availability, privacy and availability. We have made a lot of progress in the last twenty years. We have formal methods that can guarantee the accuracy of a piece of code; we have security protocols that increase the security of a particular system. And we have certain formalized notions of privacy.
Reliable AI raises the bet in two ways. All of a sudden, we’re talking about robustness and fairness—solidity, so if you mess up the input, the output doesn’t get messed up much. And we’re talking about interpretability. These are the things we never talk about when we talk about the computer.
[ad_2]
Source link