You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
It could be that publishing code promotes exploitative research, incrementally improving on last year's best, at the cost of radical explorative research that must write its own ab initio code. I don't know how serious a concern this is.
I commented on that. Barrier to entry for such incremental research becomes low, so it won't be considered innovative. Also, academics can compete on a more equal footing. We can quickly figure out which algorithms are brittle and move to deeper questions
I think the true target of my concern is not code sharing, but publication being contingent on performance superior to that of the best of a deeply-mined mode. It might take a while to build up a code base equivalent to that of the current best.
Put another way: what if the next revolutionary classifier is not a good match for tensorflow or pytorch? Would it get off the ground?
Also: fair enough, and sorry if I missed your comment.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
