AI Researchers Are Boycotting Nature’s New Machine Intelligence Journal

Researchers from Google, Facebook, IBM, MIT and Harvard have pledged not to contribute to a new journal about machine learning, citing its lack of open access.

|
May 2 2018, 12:14pm

Former executive editor of the journal  Machine Learning and an emeritus professor of computer science at Oregon State UniversityThomas Dietterich. Image: Oregon State University 

Springer Nature, the publisher of Scientific American and the venerable scientific journal Nature, intends to stride into the white-hot field of machine learning in early 2019 with a new journal called Nature Machine Intelligence.

But the community of machine learning researchers, which prides itself on publishing to open-access journals, was immediately put off by the idea of a closed-access journal that requires academic credentials to read.

Thomas Dietterich, the former executive editor of the journal Machine Learning and an emeritus professor of computer science at Oregon State University, posted a pledge not to submit, review or edit for Nature Machine Intelligence, and invited other researchers in the field to sign the pledge as well. At the time of writing, the boycott had accumulated more than 2,400 signatures by employees of Google, Facebook, IBM, Harvard, MIT and a cross-section of other prominent institutions—as well as many of the biggest names in artificial intelligence research including neural network pioneers Yann LeCun and Yoshua Bengio and Google Brain co-founder Jeff Dean.

“We write the papers, we copyedit the papers, we typeset the papers, and we review the papers,” Dietterich told Motherboard in an email. “This work is paid for by our employers. For publicly funded universities such as mine, this is a big financial transfer from the taxpayers and tuition-paying students to the publishers. Why should our employers then be expected to pay again to read the published paper?”

More so than any other academic discipline, computer science and artificial intelligence academics have long been drawn to open access publishing, which bypasses traditional gatekeepers in favor of journals like the peer-reviewed Journal of Machine Learning Research and repositories like arXiv. That commitment can also play out in the form of making new tools available for the public; Google Brain, for instance, released the machine learning framework TensorFlow as open-source software.

That commitment isn’t entirely symbolic—it also represents an effort to make computer science more inclusive, a goal with which the field has long struggled. Moustapha Cissé, a research scientist for Facebook AI Research who signed the Nature Machine Intelligence boycott, grew up in Senegal and obtained his undergraduate and masters degrees at University Gaston Berger, near Saint-Louis, where he struggled to access articles in journals his university didn’t have access to.

“I learned (the hard way) how harmful it can be to put scientific articles behind paywalls,” he wrote in an email to Motherboard. “It increases [the] technology gap which exacerbates economic inequality. Machine Learning is the most transformative technology of our time, the stakes are therefore higher.”

A spokesperson for Nature Machine Intelligence said that the journal is committed to the principles of open access, pointing to Nature’s policy of allowing authors to post preprint versions of their papers for review on platforms including arXiv, as well as to SharedIt, a program that provides authors with shareable links to published Springer Nature papers that they can freely share on social media for non-commercial purposes.

“Selective journals like Nature Machine Intelligence—which involve substantial editorial development, aim to provide high levels of author service and publish informative, accessible content beyond primary research—require investment,” said the spokesperson. “At present, we believe that the fairest way of producing these journals, which ensures their long-term sustainability as a resource for the widest possible community, is to spread these costs among many readers—instead of having them borne by a few authors.”

The conflict between Springer Nature and AI researchers highlights a broader conflict in academia. Online publishing has given rise to predatory journals that take advantage of tenure-conscious professors by publishing papers with little meaningful scrutiny, but it’s also led a generation of researchers like Dietterich to see traditional journals as unnecessary middlemen—as well as giving rise to projects like Sci-Hub and Academic Torrents, which make papers in closed-access journals publicly available in spite of copyrights.

Whether Dietterich's pledge will hurt Nature Machine Intelligence or not, this is a conflict we're going to keep seeing until there's a more meaningful shift in academic publishing.