The world is quietly being reshaped by machine learning. We no longer need to teach computers how to perform complex tasks like image recognition or text translation: instead, we build systems that let them learn how to do it themselves. "Its not magic," says Greg Corrado, a senior research scientist at Google.


6 months ago

After an intro of the switch of what training machines supposes, the author dives into deep learning as the most popular way of ML nowadays. He reminds the math has been here from the 90s (wasn't it from the 70s?) but computer power hasn't been enough until today. Than he talks how Google transitioned their products to use machine learning and open sourced tensorflow to push in that direction. Also they created a research time in Zurich about perception, natural language processing and 'machine intelligence?'. He makes a comparison of machine learning nowadays with going online straight with TCP/IP. Than he explains how he followed an article to train a network to autocomplete phrases based on the guardian articles which didn't make phrases with much sense (the fact that it built sentences is pretty good)