Monthly Archives: October 2018

It’s not over for NIPS

img_13451As you may know, over 100 Hopkins professors, researchers, grad students, and postdocs signed a letter to the machine learning conference known as “NIPS” urging a name change. The acronym is misogynist slang and a racist slur; a google or twitter search for the conference produces porn; participants at the conference have a long history of employing the acronym in ways that are hostile and demeaning. To get a taste of this one need only look back to last year’s conference, when Elon Musk is reported as having made “nips” and “tits” jokes on stage . . . to a cheering audience.

Astonishingly, the board of NIPS said no.

Those of us who support the name change must say so out loud in our departments. Silence will endorse the NIPS decision. Here are some lines you can have handy:


  • National Academies of Science, Engineering, and Medicine (NAS/NASEM) published a report this summer that 58% of women scientists have been sexually harassed and makes 15 urgent recommendations for changing the culture in STEM, including to “address gender harassment.” And NIPS doesn’t change its name?
  • Numbers don’t stand alone, and biased data don’t give accurate answers. NIPS should know better.
  • The fact that NIPS can’t use its own name for a website because it’s already claimed by a porn site is all the board needed to know.
  • Whatever you think of the acronym, the fact remains that this conference has a terrible reputation (see here and here) in terms of gender. Perception matters.
  • Whatever you think of the acronym, we just lost out on extraordinary talent among young women and their allies who want nothing to do with a field that appears to be short-sighted, tone deaf, and/or hostile.
  • A conference that sells out in less than 12 minutes is in a pretty good position to rebrand itself. What are they afraid of?
  • We are the last people who should be making it harder for girls to pursue CS, AI, ML, and the like. Just the opposite. Anything we can think of that taps into the talent of 1/2 the population is something we should do.
  • You can’t taut GirlsWhoCode one minute and then turn around and pretend that “NIPS” is gender inclusive. Or is it that they know it’s not gender inclusive, and they just don’t care?

The background noise on our campuses matters. If we speak up, even in passing, we contribute to background noise that is supportive rather than derisive or silent.

You can also sign a petition, based on the original JHU letter, protesting the decision.

There is some good news: at this year’s conferences, there will be an inclusion town hall, help with child care, gender-inclusive restrooms, a code of conduct, and mentor breakfasts for women and minorities. You can read the good, the bad, and the ugly about the Not Gonna Do It here. Just be prepared; there is some serious ugly.

Wired was right on top of this, interviewing our own Elana Fertig in “AI Researchers Fight Over Four Letters: NIPS.”  The board based its decision on analysis of a survey it sent to previous attendees at the conference.’s “NIPS AI Conference to Continue Laughing about Nipples at the Expense of Women in Tech” comments:

As a conference on what is essentially fancy statistics, you might expect them to have a better grasp on concepts like sampling bias.


Where We Stand 2018

Thank you to everyone who was able to come last Thursday, and an additional thank-you to those who helped set up!

We began with welcome remarks by Dean Toscano and an overview of the NAS report by Professor Fleming. Then, participants discussed the following recommendations in small groups:

#1 Inclusive environments and #15 The whole community is responsible

#2 Address gender harassment and #6 Support the target

#3 Move beyond legal compliance to address culture and climate

#4 transparency & accountability and #7 strong, diverse leadership

#5 Diffuse the hierarchical & dependent relationship between mentors & trainees

Details to follow!


Please share this post with colleagues

Please join other faculty, staff, and students this Thursday, October 25, from 5:30-7pm in Mudd Atrium to discuss ways to improve the climate at Homewood, across disciplines. Children are welcome.

Screen Shot 2018-10-21 at 5.02.24 PM. . . [B]y far, the greatest predictor of the occurrence of sexual harassment is the organizational climate in a school, department, or program, or across an institution. (NAS report, page 15)

. . . . Organizational climate  . . . is evaluated on three elements: (1) the perceived risk to those who report sexually harassing behavior, (2) a lack of sanctions against offenders, and (3) the perception that one’s report of sexually harassing behavior will not be taken seriously. . . .

In addition to these risk factors, there are also conditions on campus that are exacerbating the problem, including the following:

  • Insufficient attention to this topic among campus leaders—including presidents, provosts, deans, and department chairs.
  • Lack of clear policies and procedures on campus, and within departments, that make clear that all forms of sexual harassment, including gender harassment, will not be tolerated; that investigations will be taken seriously; and that there are meaningful punishments for violating the policies.
  • Minimal or merely symbolic compliance with the law without regard to whether policies actually prevent harassment and retaliation.
  • Insufficient protection for targets of sexual harassment, who often suffer undue consequences when they report sexually harassing behavior.
  • Lack of effective training on sexual harassment. While nearly all institutions offer some form of “sexual harassment training,” and often require all students, faculty, and staff to take the training, rarely is the training evaluated and revised to ensure that it has the desired effect of reducing or preventing harassment.
  • Measuring the problem of sexual harassment based on how many cases are formally reported to the institution, rather than through regular climate surveys.
  • Insufficient attention to a climate that tolerates the gender harassment form of sexual harassment, which increases the chance that other forms of sexual harassment will occur. (NAS report pages 15-16)

Columbia University collects the data: factors affecting careers of women and people of color

Screen Shot 2018-10-20 at 10.25.21 AMFrom the Chronicle this week: “What Factors Hold Back the Careers of Women and Faculty of Color? Columbia U. Went Looking for Answers”  Columbia has just published a 145-page report on how tenure-track faculty “feel about or experience key parts of academic life, like salary, workload, work-life balance, and the climate in their departments.”

“We feel like equity is a cornerstone of integrity. It’s really essential for the best scholarship to thrive,” said Maya Tolstoy, a professor in the department of earth and environmental sciences and chair of the Arts and Sciences faculty-governance committee that spearheaded the study. “We wanted to do a deep dive in the data and shine a spotlight on these issues. Now we can present it to the faculty in a transparent way and commit to action.”

. . . Tolstoy, who is also interim executive vice president of Arts and Sciences, said she was struck by how much “the smaller things” contribute to faculty feeling disenfranchised from their departments. “The added burden of committee service, the invisible labor, the belief that people don’t value their work — those kinds of things add up,” she said.

Luckily, we have a chance to talk about this sort of thing together next week, Oct 25, 5:30pm, Mudd Atrium.