3 ms·
The Rate of Progress of Deep Learning Is the Slowest in 5 Years
Today more people are working on deep learning than ever before -- around two orders of magnitude more than in 2014.
Time for something new!
- mindcrime 8y agoTime for something new! What do you suggest?
- Odenwaelder 8y ago"The slowest" by which measure?
- hsikka 8y agoI’m confused about a few things: 1. What is slowest based on? 2. Is the comment about people working in the field meant to demonstrate something about slowness or some other point?
- altairiumblue 8y agoI don't think your post makes any sense. It's on AskHN and you don't have a question. Also... - How did you define and measure progress? - How did you measure the number of people working in the field? - If true, why would any of this be a problem? - Why does a field need to constantly progress? For example, you could say that there's been even less progress in linear regression or random forest methods over the past five years - this doesn't take away from their usefulness in certain applications. - Why would the assumed slow rate of progress mean that DL needs to be replaced? - If you find more suitable solutions, you're always free to implement them in place of DL.