|Peter Norvig - As We May Program|
|Written by Sue Gee|
|Sunday, 19 May 2019|
In the spirit of sharing interesting items that we stumble across, here is a video of a lecture delivered by Peter Norvig as the latest Microsoft Research Distinguished Lectures Series.
Peter Norvig is a seasoned speaker and had given variations on this lecture multiple times prior to this, so apologies if you may have already encountered it. If not it's worth viewing for anyone with an interest in computer programming and/or machine learning.
Currently Director of Research at Google, when he joined Google in 2001 Peter Norvig's title was Director of Machine Learning, but at that time he didn't envision how important that topic would become in the space of almost two decades.
Of course, Norvig himself is responsible for our exponentially expanding interest in Artificial Intelligence and Machine Learning in particular. He we co-teacher of the 2011 Introduction to AI Class that signed up 160,000 students, helping to kick off the MOOC phenomenon that continues to attract thousands of learners to Computer Science and AI in particular. Recently he turned up in our news stream introducing the free Google Machine Learning Crash Course, a starting point for anyone to get into this area.
The general theme of this lecture is the future of programming and computer science in the context of the progress being made in AI and ML. As the description put it:
Innovations in machine learning are changing our perception of what is possible to do with a computer. But how will machine learning change the way we program, the tools we use, and the mix of tasks done by expert programmers, novice programmers, and non-programmers? This talk examines some possible futures.
As acknowledged in the opening slide, the title of his talk is inspired by a Vannevar Bush 1945 essay entitled "As We May Think" and the beginning of the lecture looks back to the emergence of computer programming, when software was largely based in mathematics:
Looking to the past Norvig reminds of of the contrast between the tiny power to size ratio of EDVAC with today's smartphones, raising a laugh from the audience with this slide:
Looking to the future Norvig speculates on how our interaction with technology will evolve as intelligent agents become increasingly prevailing in our lives and how Computer Science is gradually becoming an empirical science:
The lecture concludes on a cautionary note about the dangers inherent in recommendation systems. By building systems that measure what we want in terms of click rate we effectively spread a virus that promotes short term wants over society's real needs:
This is a very valid point and one that needs to be heeded. Thank you, Peter Norvig.
or email your comment to: firstname.lastname@example.org
|Last Updated ( Sunday, 19 May 2019 )|