AI Progress Debate Meltdown

Another ridiculously overhyped deep learning paper. These GPU junkies should learn a little about the history of AI before posting this crap on arXiv. You're just jealous that after decades of near-stagnation, the field is finally making rapid progress, and you aren't getting all the credit. Generating fake faces and beating decades-old games doesn't count as rapid progress to me. Where's the common sense reasoning? Stop shifting the goal posts! You said superhuman Go was a grand challenge for AI, and now it's been done, no thanks to you. If you want to have a serious debate about this, why don't you submit a short paper to the ICML 2018 Debates workshop instead of yelling like a loser?

A five-panel argument shows two men angrily debating the progress of artificial intelligence, from deep learning papers and fake faces to superhuman Go and common-sense reasoning. The joke is that an academic disagreement escalates into increasingly personal accusations and shouting, making it useful as a reaction to someone moving the goalposts or overreacting in a technical debate.

• @profunctor_io / 737 2018-03-30