Update about.md

_pages/about.md

@@ -28,7 +28,8 @@ and safeguarding
 AI for equitable and robust deployment in high-stakes environments. Towards this goal, I study both algorithmic foundations and empirical methodologies, centered on:
 
 
-- **<span style="color:darkred;">Sustainable AI:</span>** Developing principled methodologies to measure and minimize AI's resource usage and lifecyle environmental footprint (**[Communications of the ACM](https://arxiv.org/abs/2304.03271), [e-Energy'24a](https://arxiv.org/abs/2405.17469), [e-Energy'24b](https://arxiv.org/abs/2311.03615), [ASPLOS'24](https://dl.acm.org/doi/abs/10.1145/3620665.3640374), [SIGMETRICS'22a](https://arxiv.org/abs/2111.01203)**)
+- **<span style="color:darkred;">Sustainable AI:</span>** Developing principled methodologies to measure and minimize AI's resource usage and lifecyle environmental footprint (**[Communications of the ACM](https://arxiv.org/abs/2304.03271), [e-Energy'24a](https://arxiv.org/abs/2405.17469), [e-Energy'24b](https://arxiv.org/abs/2311.03615), [ASPLOS'24](https://dl.acm.org/doi/abs/10.1145/3620665.3640374), [SIGMETRICS'22a](https://arxiv.org/abs/2111.01203)，
+[OECD AI Policy Observatory <span style="font-weight:lighter;color:darkred;">(perspective article)</span>](https://oecd.ai/en/wonk/how-much-water-does-ai-consume)**)
 
 - **<span style="color:darkred;">Safe decision-making:</span>** Robustifying machine learning predictions in highly dyanmic, uncertain, and/or adversarial environments such as renewable-powered computing systems (**[SIGEMETRICS'24](https://arxiv.org/abs/2401.04340), [NeurIPS'23a](https://arxiv.org/abs/2311.01568), [NeurIPS'23b](https://arxiv.org/abs/2310.20098), [ICML'23](https://arxiv.org/abs/2306.00172), [ICLR'24](https://openreview.net/pdf?id=e2YOVTenU9), [SIGMETRICS'22b](https://arxiv.org/abs/2204.08572)**)