AI, Gatekeepers & New Rules for Solving the World’s Biggest Problems — BigThinking.io
Did you know that globally only 22% of AI professionals are women? Most business leaders agree that artificial intelligence (AI) will play a key role in shaping the future of work…
Did you know that only 22% of Artificial Intelligence (AI) professionals globally are female? Most business leaders agree that artificial intelligence (AI) will play a key role in shaping the future of work and solving some of the world’s most critical problems. Leaders in AI research are widely considered “the best and brightest” (also see “top talent”), yet many seem to be okay with women and people representing diverse cultures and socioeconomic backgrounds being grossly absent from the conversations that decide what and how important problems are solved with AI. As of June 2019, Google’s AI research page listed 641 people working on machine intelligence, only 10% were women. As of June 2019, Facebook’s AI research group included 15% of women ( source: Wired).
Why are Diverse AI Teams Important?
The people most impacted by broken systems are often left out of the discussion regarding how AI can solve the world’s pressing issues; they include women, LGBTQ+, low-socioeconomic status communities and ethnic & racial minorities. For example, in 2018, with unprecedented access to emerging technology, data, and talent, Amazon imagined, developed and then shuttered an intelligent hiring tool when it was revealed that the system penalized women job candidates. Much of the press stated that the AI was not ready for prime time because it (it = the AI) showed bias against women. Very little press was dedicated to evaluating the root cause of the bias or the environment that empowered the AI team to use biased datasets to train an AI that would ultimately penalize job candidates that identify as women. There are too many similar examples, with public statements by the world’s biggest leaders promising to deploy the same teams of the “best & brightest” to research technical fixes to AI’s pesky bias problem. Bless their hearts.
The “expertise trap” serves as a blind spot for many brilliant people that are often rewarded and recognized for their ability to maintain careers in highly technical fields. Your experience is well earned, but it is also your responsibility to consider that you don’t know everything you need to know to solve every problem. Given the global impact of increasingly intelligent systems, it is the responsibility of all leaders to reevaluate and challenge our assumptions regarding who gets to participate in the innovation process. It’s time to open the gates
New Rules for Solving the World’s Biggest Problems
- Who decides what’s important? It is our responsibility to require diversity in the voices that determine “what’s important” and that highlight opportunities to solve new problems with AI. With broader perspective, we expand the space of ideas that can benefit from effective use of machine intelligence.
- Who understands the social impact? It is our responsibility to understand the social impact of intelligent systems and interconnected systems and domains. The people most critically impacted by these systems MUST be involved in a meaningful way. AI solutions cannot be designed without the people that understand the domain & its social impact; these are your subject matter experts.
- Who are your valued team players? It is our responsibility to understand the importance and value of cross-disciplinary AI teams. This is important work and everyone can add value, equally. Framing the discipline as “rocket science” that can only be performed by a privileged few has only served the egos of the people currently working in the space & leading the conversations. Certainly, it is science, but it’s also math and business and art and empathy and critical thinking and change management and communication … clearly, engineering alone won’t create a “good AI” solution. You will need more people.
The best AI built on top of ineffective, broken and biased systems will only accelerate and amplify the impact of ineffective, broken and biased systems . The worst attempts at leveraging AI are often attributed to the technology not being ready for “prime time”; it’s not the technology that isn’t ready, it’s the people that aren’t ready. It’s beyond time to challenge our thinking regarding who can and should participate in solving the world’s biggest problems.
Originally published at https://bigthinking.io on December 2, 2019.