The Struggle is SO Real: Pervasive Gender Bias in STEM
New research about women in STEM finds that over 40% of women leave fulltime STEM positions a year after becoming a parent. Researchers point to pervasive gender bias and structural problems within the field, but say that the problem goes beyond retention. Even women who stay employed full-time in STEM face gender bias in promotion and performance evaluations. Male managers are more likely to be promoted, mentored, and rated as more competent than women with the exact same resume. Another obstacle is lack of mentors. Sheryl Sandberg has called out senior supervisors who refuse to mentor or even meet alone with female colleagues: “If you’re going to have dinners, then invite women and men and behave appropriately. If you’re not going to have dinners with women, then don’t have them with men.” Some analysts have dubbed it the ‘Pence Protocol’ since the VP also refuses to dine alone with women. Despite study after study after study, many men still deny or downplay the discrimination.
Think that this situation is just terrible but that you’re not biased? Maybe you should test your own implicit biases about gender and work to find out.
Researchers Catfish NATO Troops
NATO researchers recently conducted an experiment to see if they could engage with and influence the behavior of NATO troops using fake Instagram and Facebook accounts. They could, and they did—easily. NATO’s Strategic Communications Center of Excellence just released the results of the covert experiment in a report focusing on ‘cognitive security challenges’. The report details how the researchers, using only social media and open-source records, were able to identify individual service members, track troop movements, and convince soldiers to disobey orders. The report highlights both the ease with which the researchers were able to infiltrate the troops as well as the troops’ overall vulnerability to disinformation. The exercise (and report) are part of a larger efforts aimed at combating ‘next-generation’ security threats. Report authors recommend additional training. Meanwhile, Russian officials have already banned troops from using mobile devices and social media—perhaps due to their own intimate knowledge of just how ridiculously effective disinformation campaigns can be.
Genetic Surveillance: China Harvests DNA from Minorities & Uses It to Track Them
Authorities in China’s western Xinjiang region have secretly built a DNA database to track the region’s Uighur muslim population. Chinese officials forcibly collected blood samples, facial scans, and voice recordings from millions of Uighurs during bogus ‘health screenings’. Coerced collection of DNA is the latest operation in Beijing’s strategy of brutal persecution of its ethnic minority population. More than a million Uighurs have been detained in reeducation camps, and authorities are believed to use the DNA data to locate individuals critical of the government and to check attendance at reeducation camps. This latest revelation has provoked a crisis in the scientific community since China’s DNA harvesting relied on American equipment and researchers. China’s use of openly-shared data to oppress citizens highlights the complexities of dual-use technology in foreign policy and national security. In Australia, lawmakers have recently reversed restrictions on technology with military potential. This isn’t the first time that China has abused publicly available research and consumer technology. Earlier this week, a data leak accidentally exposed in shocking detail the extent of the state’s surveillance capabilities.
AI News Anchors Should Legitimately Terrify You
China’s official news agency has announced its latest AI news anchor, Xin Xiaomeng. Xin will make her official debut with China’s other two AI news anchors later in March. In fewer than 4 months, the AI anchors have produced over 3,000 news reports and have delivered 167 hours of broadcast coverage. China plans to use the anchors to manage social media platforms and to write and deliver news stories. In the US, OpenAI (a non-profit AI research firm) has unveiled an algorithm that can convincingly imitate the writing style of sample texts. Really convincingly. OpenAI has so far refused to release the system for fear of misuse. Initially, researchers thought that AI could be used to combat the spread of fake news. Google and Facebook both launched initiatives using AI to police hate speech. However, researchers now warn that the very same technology behind Xinhua’s AI news anchors could just as easily be used to mass-produce fake news. Perhaps China should go ahead and program the anchors to just communicate in Newspeak.