layout: single title: “Springtime for machine learning” tags: tech machine-learning google microsoft facebook research
As a difficult year draws to a close, I’d like to end 2016 on a hopeful note. Machine learning has broken through to the mainstream. In a real sense, artificial intelligence is here to stay. This is a bit surprising because AI has seen half a dozen booms and busts since the 1950’s. The cycle is so frequent that the term AI winter was coined many years before Jackie Fenn introduced the Gartner hype cycle in 1995.
Here are some favorite pieces from the past year:
-
Over the summer, Steven Levy at Backchannel detailed how Google is retraining its engineers in a bid to move Alphabet from being a search company to a mobile company to an AI-first company.
-
In March, Jeff Atwood published a great blog post about AlphaGo and how the last decade of GPU advancements has been critical to the rise of practical machine learning / deep learning.
-
Daniel Terdiman wrote an in-depth Fast Company article about how Yann LeCun is building the Facebook AI Research (FAIR) program in collaboration with the engineering-oriented Facebook Applied Machine Learning (AML) team led by Joaquin Candela.
-
In September, Microsoft announced a major reorganization where Microsoft Research (founded 25 years ago) combined with AI staff previously distributed throughout the company to form the Microsoft AI and Research Group with 5,000 computer scientists and engineers.
-
Just last week, the New York Times published a long but very worthwhile Sunday Magazine article by Gideon Lewis-Kraus about how Google has moved AI/ML out of the lab and into the hands of half a billion users per month. The article starts with multilingual translation but goes on to talk about image recognition, artificial vision, and many other things.
-
Finally, I encourage everyone to watch this excellent slideshow by Frank Chen at Andreessen Horowitz. It provides a nice history of AI and excellent overview of machine learning and deep learning.