It’s time to allow researchers to submit manuscripts to multiple journals

For example, in continual deep learning people often use time datasets in which they use a small amount of memory and incrementally learn classes and the algorithms cannot work for other incremental learning distributions. It's been very hard to publish work that instead works well for arbitrary multiple distributions, eliminates the memory constraint (which doesn't matter in the real world mostly), and shows it scales to real datasets. We have been able to get huge reductions in training time with no loss in predictive ability, but can't seem to get any of these papers published because the community says it is too unorthodox. It is far more efficient than periodically retraining as done in industry, which is what industry folks always tell me is the application they want from continual learning.

The confusing thing is that when I give talks or serve on panels I always have folks thank me and tell me they think this is the right direction and it was inspiring.

In my field the review system is way overtaxed with too many inexperienced people who struggle with having a broad perspective, so I think submitting to more venues would probably make things worse.

A less spirited method is to let traditional publishing slowly die by adding review features to highly viable alternatives like arXiv, PubMed and others which will then replace it. I hate academic publishers, they are vultures that serve no purpose. I hope they all go out of business or somehow pay big :)

One thing that struck me as a researcher back in the day was that it was much easier to get people reading things that I put on my blog than it was getting scientists to read my papers. Basically as a researcher, you are basically engaging in 17th century SEO by writing papers and getting your peers to pay attention to this. We use email these days but otherwise the process hasn't changed a lot.

This is weird considering we now have a lot more tools. Imagine if hacker news worked that way. It wouldn't work. The best kind of endorsement for a paper is not random anonymous peer reviewers giving their thumbs up or down as to the inclusion of a digital only journal that nobody ever reads cover to cover. Instead it is other researchers citing your work. This is in fact part of how most academic performance is measured. The goal is to get your peers being aware of the existence of your work, get them to spend time reading or learning about it, and then getting them to engage with it by citing, criticizing, or adding to it.

The whole business of a journal publication is just SEO. You get some renowned journal to include you and maybe people will bother reading it and maybe refer your work if they like it. A citation is just a glorified like. The real goal is to get people to read and "like" your work. It's also self re-enforcing: the more people cite your work, the more people will read it and thus cite it. But it's a super inefficient process. From idea to citations happening can take years.

Conferences and workshops are where scientists meet up and discuss their work. That's where your peers are. I always enjoyed the smaller workshops. Get some smart people in a room and beautiful things happen. The academic world is basically a social network. With likes and everything. But minus all the modern tools that make other social networks work so well. There's some room for improvement.