13 ms·
Hate to be that person, but - If you can go from zero to SOTA from a web tutorial in anything less than, I don't know, a year, then the field is severely under
by zxcvvcxz 8y ago
Hate to be that person, but -
If you can go from zero to SOTA from a web tutorial in anything less than, I don't know, a year, then the field is severely underdeveloped.
But let that be an opportunity: what it really means is that the field is quickly growing and there aren't well-established experts or leaders.
- laGrenouille 8y agoLooking at the material here, I think you may be under estimating what is meant by "zero" here. To fully follow these notes you would need to have significant research-level knowledge of general machine learning techniques, with a good deal of specific experience working with neural networks (both in theory and application). I think someone with a general graduate-level background in most fields of ML or computer science could acquire a working knowledge of the SOTA in a particular associated sub-topic after reading through a half dozen or so research papers and code. That doesn't seem unreasonable or surprising.
- backpropaganda 8y agoHe's also under estimating what it means to be a hero. Most of the experiments here are on Cartpole which is a very basic environment (not even MNIST level of complexity). Most papers in RL use Atari, which requires some amount of engineering that this repo does not have.
- make3 8y agocome on man, the title is obviously click bait; you can only begin to build an understanding of a field with a single tutorial, even if it's super in depth
- tartrate 8y agoUnderstanding and making use of E=mc^2 takes less time than it took to research it.
- __bee 8y agoHate to be that person, but - This is not the spirit that some of us want to see in HN. Efforts are always appreciated. The field is improving incrementally, in a very fast way.