I feel sorry for the postdoc here. You have a seemingly decent paper, but the press release really doesn’t do it justice. For reference, the Tweet about the pre

I feel sorry for the postdoc here. You have a seemingly decent paper, but the press release really doesn’t do it justice. For reference, the Tweet about the press release https://twitter.com/princeton/status/1296779082663964673?s=21 https://twitter.com/princeton/status/1296779082663964673?s=2... vs the paper https://sci-hub.tw/downloads/2020-08-10/f1/10.1038@s41562-020-0924-8.pdf https://sci-hub.tw/downloads/2020-08-10/f1/10.1038@s41562-02... The press release DOES have a very “machine learning ‘discovers’ trivial thing people already knew” vibe. The actual paper doesn’t - it’s not trying to “rediscover” anything, it’s just building on existing work, there’s none of the arrogance in the press release that’s triggering everyone on Twitter. The press release taints the actual research, which is a sadly common pattern, and I think a fair bit of that lies with the writers of press releases, not just the readers.