NeurIPS 2020 Highlights Ethics Clashes and Corporate Influence in AI
The NeurIPS 2020 conference highlights major ethical conflicts as researchers protest Google's firing of Timnit Gebru and scrutinize sponsors like Huawei. Meanwhile, workshops emphasize the urgent need to shift AI development away from corporate dominance toward empowering marginalized communities.
Organizers of NeurIPS, the world's largest AI research conference, award Best Paper honors to three studies, including one detailing OpenAI's GPT-3 language model. The conference opens under a cloud of controversy as AI researchers refuse to review Google's papers to protest the firing of Ethical AI team co-lead Timnit Gebru. Google employees describe her dismissal as unprecedented research censorship that raises serious concerns about corporate influence in the AI field.
Tensions between corporate interests and human rights dominate various workshops throughout the week. Participants at the Muslim in AI workshop examine GPT-3's anti-Muslim bias and the use of AI for surveillance in China, drawing uncomfortable attention to platinum sponsor Huawei. In response to these ethical concerns, a NeurIPS spokesperson reveals that the conference is forming a new sponsorship committee to establish proper vetting policies for accepting future sponsors.
The Resistance AI workshop pushes back against superficial corporate ethics by showcasing research that compares Big Tech's funding tactics to those historically used by Big Tobacco. Organizers express frustration with the limitations of "AI for good" initiatives, which they say companies easily co-opt as ethics-washing. They advocate instead for an intersectional approach to AI development that actively returns power to marginalized people and avoids systems of oppression.