
Episode 13: Jonathan Frankle, MIT, on the lottery ticket hypothesis and the science of deep learning
Generally Intelligent
00:00
Publication Doesn't Have to Be Significant
We're essentially paying forward some poor treatment that no junior researches were receiving at the mila. We can do better. Publication doesn't have to be suffering y i look tat i lok ta awit exactly how you put it does not mant hal to mean its wos significante. An cand just be true. The reviewing process is just to arbitrate truth. Yes, we want to make sure that things are well justified enough and that they're probably right. But why should we strive for having some arbitrary bar? Why not just accept things that are true and let people get them out there and then let them move forwardi wolt diceminate knowledge.
Transcript
Play full episode