Fresh Grad vs Senior Dev

FRESH GRAD SENIOR DEV I ♥ NOOOOOOOO!!!!! YOU SHOULD USE DEEP NEURAL NETWORKS IN THIS PROBLEM!! ONLY 200+ HOURS OF TRAINING ON BIG CLUSTER OF TPU'S CAN GRANT YOU DECENT PERFORMANCE!!!!! OMG YOUR MODEL GIVES 3% LESS ACCURACY NOOOOO!!!!! haha TF-IDF + LOGISTIC REGRESSION GO BRRRRRRRRRR

A frantic fresh graduate demands deep neural networks and massive TPU training, while a calm senior developer uses TF-IDF with logistic regression. The joke mocks overengineering and the tendency to prefer fashionable deep learning over simple methods that work well. EN: A frantic fresh graduate demands deep neural networks and massive TPU training, while a calm senior developer uses TF-IDF with logistic regression. The joke mocks overengineering and the tendency to prefer fashionable deep learning over simple methods that work well.

• @profunctor_io / 4457 2020-04-20