[TheBiv]

The study cited is a 3 page PDF that has absolutely zero scientific weight.

https://arxiv.org/pdf/1712.00676.pdf

[tananaev]

I don't think there can be any actual science behind predictions like this. Someone just wants more funding for AI research, I guess.

[igorgue]

I cannot believe is only 3 pages do you really need 3 authors to do that?

At least this kind of academia wont be replaced because their work is worthless, thus we have no interest into automating it.

[viraptor]

It doesn't matter how many pages are there in a paper. It's the content that matters. Math papers have the best chance of standing out here due to good abstractions: https://mathoverflow.net/questions/7330/which-math-paper-max...

If you discovered something amazing that takes one page to describe, why write a book about it? Even if it took 80 people in the team.

PS. exactly this kind of work (crappy paper) has been automated: http://thatsmathematics.com/mathgen/ ;-)

[revelation]

Some of the best papers of past were only 2 or 3 pages. Certainly not this one though, that's not even a paper, that's just an extended online comment, letter to the editor, journal preface.