Recently, a speaker at NeurIPS, the renowned AI conference, has come under fire. The controversy arose not from her views on AI, but from an insensitive reference to a Chinese student during her presentation. Rosalind Picard, a professor at MIT Media Lab, presented her keynote speech on ‘Optimizing What Matters Most.’ She included a slide that quoted a ‘Chinese student, now expelled from a top university,’ blaming the lack of ethical education for his misuse of AI. Picard added a note stating that most Chinese people she knows are honest and morally upright.
The slide sparked a reaction when Jiao Sun, a Google DeepMind scientist, shared a picture of it, commenting on the challenge of eliminating racial bias from humans compared to Language Learning Models (LLMs). Yuandong Tian, from Meta, echoed Sun’s sentiments, questioning how such explicit racial bias could occur at NeurIPS.
During a Q&A session, an attendee pointed out that Picard’s reference to nationality was unnecessary and potentially offensive, suggesting she consider removing it in future presentations. Picard seemed to agree with the suggestion.
NeurIPS organizers issued an apology after the incident, asserting that such comments do not align with their code of conduct and that they are addressing the issue with Picard directly. They emphasized NeurIPS’ commitment to diversity, inclusivity, and equal treatment for all.
Picard also issued a statement to express her regret for including the student’s nationality in her presentation. She acknowledged her mistake and apologized for any harm caused, promising to learn from the experience and seeking suggestions to make amends to the affected community.
