Margaret Mitchell: Some may have noticed that there is a relatively large amount of *women* from HF who speak in media. As scientists & nerds, we are ofc sharing our experiences and thoughts before+after. It's been a unique opportunity to examine the following research question: 2/

Have a lot to say; I'll start from punchline: -There are many women @huggingface.bsky.social who have had the opportunity & privilege to speak to press about AI. -We have noticed gendered patterns. So, we put together a resource for journalists on the patterns. www.linkedin.com/pulse/rules-... 🧵

Rules of [Press] Engagementwww.linkedin.com How reporters should talk about women in tech, especially women in AI.

Some may have noticed that there is a relatively large amount of *women* from HF who speak in media. As scientists & nerds, we are ofc sharing our experiences and thoughts before+after. It's been a unique opportunity to examine the following research question: 2/

"What are the patterns in how journalists talk to, and about, women in AI?" 3/

We've discovered that, compared to our male peers, there is a disproportionate focus from press on: - Our ages - Our motherhood - Our physical appearance or behaviors - Our failures - What AI "gossip" we can provide rather than technical work 4/

After the most recent double-whammy, in which I was described in one news outlet as "flirtatious" and cagey about my age, and @sashamtl.bsky.social was described as a mother balancing work and life, we decided we'd put together some "good practices". 5/ www.linkedin.com/pulse/rules-...

Rules of [Press] Engagementwww.linkedin.com How reporters should talk about women in tech, especially women in AI.

With this doc, we have the following goals: 1. To help lessen the tendency for press to perpetuate/reify conceptions of women that harm our ability in tech to inspire, recruit and retain women. 6/

While we can have great initiatives like WiML, these exhausting efforts are massively hampered by press narratives that discourage women from identifying with Machine Learning as something they can and should be a part of. 7/ wimlworkshop.org

Women in Machine Learningwimlworkshop.org

2. To help create a culture in press where technical women who speak with journalists don't have to undergo feeling alienated, demeaned, regularly misrepresented, or violated. 8/

Post