TECHNOLOGY

Transformation of Computing | MIT Technology Review


Wing himself never wanted to study computer science. In the mid-1970s, he entered MIT to do electrical engineering, inspired by his father, a professor in that field. When he discovers his interest in computer science, he calls her and asks if it’s a passing fad. After all, there were no textbooks in the field. He assured her that it was not. The Wing Major has changed and never looked back.

Former Corporate Vice President of Microsoft Research and now Executive Vice President for Research at Columbia University, Wing is a leader in the promotion of data science in multiple disciplines.

Anil Anantaswamy Wink was recently asked about his ambitious agenda to promote “trusted AI”, one of 10 research challenges he identified in an effort to make AI systems more fair and less biased.

Q: Would you say that there is a change in the method of calculation?

Answer: Absolutely. Moore’s law has taken us far. We knew we were going to hit the ceiling for Moore’s law, [so] Parallel computing is predominant. But the phase shift was cloud computing. The original distributed file system was a kind of baby cloud computing, where your files were not localized to your machine; They were somewhere else on the server. Cloud computing takes it and makes it even more so, where you don’t have the data; The count is not near you.

About the next transfer information. For the longest time, we’ve been stuck on the cycle so that things work faster — processors, CPUs, GPUs, and more parallel servers. We ignore the information part. Now we need to fix the data.

Q: This is the domain of data science. How would you define it? What are the challenges of using data?

A: I have a very short definition. Data science is the study of extracting values ​​from data.

You can’t give me a bunch of raw data and I press a button and the value comes out. It starts with collecting, processing, storing, managing, analyzing and visualizing data and then explaining the results. I call it the data life cycle. Every step of that cycle is very useful.

Q: When you are using big data, there are often concerns about privacy, security, fairness and bias. How does one deal with these issues, especially in AI?

A: I’m promoting this new research agenda. I call it trusted AI, we are inspired by decades of progress in trusted computing. By fidelity we usually mean security, reliability, availability, privacy and usability. We have improved a lot in the last two decades. We have formal procedures that can ensure the accuracy of a portion of the code; We have security protocols that enhance the security of a particular system. And we have some idea of ​​privacy that is formally created.

Reliable AI promotes progress in two ways. Suddenly, we’re talking about robustness and fairness – perseverance means if you bother the input, the output doesn’t bother too much. And we’re talking about explanation. These are things we never talked about when we talked about computing.



Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button