My take on Ali Rahimi's "Test of Time" award talk at NIPS. Ali gave an entertaining and well-delivered talk. But I fundamentally disagree with the message. The main message was, in essence, that...
"It's not called alchemy, it's called engineering." Interesting response from
and discussion of Ali Rahmi's test of time talk
's take on
talk. I didn't take it as theory vs. DL. Theory not only tool for rigor. NN progress is awesome, but rigor in DL experiments not uniform: simple experiments, stronger baselines, deeper understanding help everyone
Shots fired -
's much discussed talk:
One of the big talking points at #NIPS2017 so far is
's comparison of machine learning to alchemy--and the response by
Personally, I think Ali made an important point that certainly isn't well understood by the public.
Ali Rahimi's talk at NIPS and his comparison of ML to Alchemy.
Well delivered and spot on in many aspects.
has a different opinion: