A Recap of Women in Tech Initiative Director Jill Finlayson’s Grace Hopper Celebration 2020 Session
Increasingly, AI-enabled tools are used in every facet of the hiring process. Companies can’t keep up with turnover rates and hiring demands. For better or worse, AI is here to stay. And those that participate in the hiring process need to do their part to make it for the better. We need diverse teams in tech to make sure AI works for everyone.
Kari Paul, The Guardian, San Francisco, 16 Apr 2019
‘Disastrous’ lack of diversity in AI industry perpetuates bias, study finds [AI Now Institute] Report says an overwhelmingly white and male field has reached ‘a moment of reckoning’ over discriminatory systems
So to help you find a job in tech, and/or to hire a more diverse and talented team for your company — let’s level-set on where and when you will encounter AI in the recruiting process, and how you might prepare.
STEP 1: Finding (& listing) a job in tech:
AI can help hiring managers write better job descriptions and job applicants hone their resume, but if recruiters click “optimize” when they post a job, AI can amplify bias (by measuring success by clicks rather than reaching the targeted qualified applicant pool).
In the future, AI may surface jobs you are a good match for, and even help identify skills gaps and trainings to move up the career ladder. But for now, it falls on you to find the jobs you are qualified for and to architect your own career path.
TIP: Try more targeted job boards aiding intentionally inclusive companies. You may want to clear your cookies and submit applications from an incognito window to avoid AI from limiting your job discovery.
Choosing a company: Glassdoor, in addition to job listings, enables you to search crowd-sourced job titles, salaries, and company culture, including a new feature for feedback on gender equity and diversity and inclusion efforts.
STEP 2: Applying
When you find a job, you will likely meet a chatbot like Miya, Ideal, Brazen, GoHire, Paradox, and Wade and Wendy. Chatbots can be quite useful, collecting information, answering frequently asked questions, scheduling interviews with humans, chatting by SMS, email, social media, but they may also be asking screening questions.
TIP: Assume you are speaking to a chatbot and it could be filtering your resume in, or out. Answer carefully and assume the information is being kept with your application, just in case. Also, be aware that a company may look at your publicly available information (Fama.io) and could screen you out based on toxic behavior (hate speech, pending litigation), so Google yourself and clean up any public information you would not want a company to see. Another concern here is the potential for conflating your identity with someone else with the same name.
STEP 3: Screening and Filtering
AI has the potential to anonymize applications and enable people to be found based on having the skills to do the job. This could be a boon to equity, given that hiring discrimination against black Americans hasn’t declined in 25 years, and males and white-sounding names are significantly more likely to get an interview than identical resumes with names that sound Asian, Black, Latinx, or female.
However, if we don’t control for biased criteria and datasets, AI will detect patterns and may replicate or amplify bias. Amazon, using datasets that skewed heavily male, ended up scrapping their hiring tool, because it started discriminating against women, and the AI could not “unlearn” the bias it developed. AI, on the bright side, could encourage hiring managers to revise their requirements by enabling them to see how the talent pool can be expanded by reducing the number of years of experience required, or by selecting a masters instead of a MBA — which produces both a more inclusive and qualified pool of applicants.
TIP: Customizing your resume to match the minimum required skills will be essential to make it through the filters. And whether you make it through the filters or not, networking and warm referrals remain essential for getting a foot in the door. Visit the social media sites of the companies you are interested in to discover and take part in their events and recruiting programs.
STEP 4: Evaluating & Testing
Structured interviews are far superior to unstructured interviews, and AI can provide everyone with the same interview questions. It also offers convenience and efficiency for scheduling early rounds of interviews. You may encounter personality assessments for essential skills like leadership, collaboration, etc. Games and testing for skills could potentially replace the resume with a much better assessment of skills and ability to do the job, regardless of degree or years of experience.
However, it is early days, and the technology is not yet ensuring an equitable experience for all and there is a lack of independent auditing for equity. Most controversial is the use of facial recognition software to assess a candidate’s potential and fit for a position. “Facial action units” make up to 29 percent of a candidate’s score based on “intonation,” “inflection” and “emotions” to determine social intelligence (interpersonal skills), communication skills, personality traits, and overall job aptitude. Gaming systems are not free from bias either, with greater barriers for older workers, non-native English speakers, and others with personal (and preferably private) limitations.
TIP: Practice, persist, and be professional — do mock interviews, practice coding problems, and prepare your space for the interview: put your phone away (but have it nearby in case Wi-Fi goes out and you have to tether your phone), close other applications/notifications, find a quiet space, consider what is in the background, ensure good lighting (ring lights are pretty cheap actually), practice looking at the camera, and dress appropriately for the company. The best video quality has replaced the best suit for establishing your credibility as a candidate, which is not fair, but true.
STEP 5: Find your community
You are not alone in finding your way in this new landscape. Join a community for support, camaraderie, resources, and networking. As I share this information with the Grace Hopper Community today, there are also many other associations and groups to join. Here are a few: NSBE-National Society of Black Engineers, SHPE-Society of Hispanic Professional Engineers, SWE-Society of Women Engineers, AISES-American Indian Science and Engineering Society, and IEEE.
STEP 6: Make AI Better and More Inclusive
Invite AI experts to speak at your event: Consult the Open Directory of Women in AI Ethics
Get involved in AI Engineering and Ethics and strengthen diversity and inclusion:
AI4All opens doors to artificial intelligence for emerging talent through education and mentorship
The Algorithmic Justice League’s mission is to raise public awareness about the impacts of AI, equip advocates with empirical research to bolster campaigns, build the voice and choice of most impacted communities, and galvanize researchers, policymakers, and industry practitioners to mitigate AI bias and harms.
Alliance for AI Africa’s sharpest innovators coming together to not only generate wealth through technology advantage but to slow down the unfolding bias and unfairness from Africa’s exclusion from the AI revolution. (HT Aretha Mare, TechWomen)
A Crowdsourced Algorithmic Bill of Rights drafted by AI experts highlight 10 things we should demand: Transparency, Consent, Freedom from Bias, Feedback Mechanism, Portability, Redress, Algorithmic Literacy, Independent Oversight, Federal and Global Governance
At the University of California: Alliance for Inclusive AI, AFOG-Algorithmic Fairness and Opacity Working Group, Berkeley Center for Law and Technology, Berkeley Law: Samuelson Law, Technology, & Public Policy Clinic, Center for Effective Global Action: CEGA, CHAI-Center for Human-Compatible AI, Citizen Clinic, CITRIS Policy Lab, CTSP-Center for Technology, Society & Policy, Human Rights Center, and GEESE
Expert Seminar Series: Leading Trends in Human Resources including a session on The New HR — Employing Equitable AI by Brandie Nonnecke.
Research and Guidance on Developing/Deploying AI Ethically:
Mitigating bias in algorithmic hiring: evaluating claims and practices and FATE: Fairness, Accountability, Transparency, and Ethics in AI (Microsoft)
Mitigating Bias in Artificial Intelligence Playbook, Berkeley Haas Center for Equity, Gender, and Leadership (EGAL)
Could Ratings Systems Promote Responsible AI (Brookings)
World Economic Forum WEF AI for HR guide (in development)