CELEBRATING HMI WORKSHOP

Photo by George Prentzas on Unsplash

On 13 December 2023, HMI ran a workshop exploring AI development and the future of integrated AI.

The workshop opened with a keynote address by Xiao-Li Meng and James Bailie from Harvard University titled ‘Privacy, Data Privacy and Differential Privacy’. Following the keynote address was a session on HMI Research Highlights, touching on the four years of research HMI has done as AI and human morality continue to intertwine. Presentations were made by Katherine Bode, Katrina Sluis, Ziyu Chen, Lydia Lucchesi, Rahul Shome, Nick Schuster, Lexing Xie and Sujatha Raman.

The workshop continued with presentations of eight selected seed grants that will continue under HMI. The seed grant recipients’ topics ranged from the integration of computing with climate change and weather patterns, to the fusion of AI in the creatives. The recipients of the seed grants are as follow:

Human-Machine Aesthetics by Katherine Bode and Sabrina Bleecker Caldwell

Virtues of Robot Inaction: Towards Theories of Automated Reasoning of Inaction in Human Contexts by Rahul Shome

Privacy-Preserving Perception for Robotics by Yunzhong Hou, Dylan Campbell, Rahul Shome and Michael Barnes

Chasing Storms with AI-Enhanced DAS, Seismic and Infra-sound Arrays by Rhys Hawkins, Voon Hui Lai and  Chengxin Jiang

Ways of Seeing Datasets: Toward Socio-Cultural Understanding of Machine Vision by Katrina Sluis and Liang Zheng

ANU Bushfire Smoke Dataset by Gao Zhu, Nick Wilson, Weihao Li and Nick Barnes

Cross-Domain Sampling Methods for Greener High-Performance Computing by Dr Amanda Parker and FHEA

Advancing Machine Learning-Assisted Modelling of Sea Ice Dynamics by Quanling Deng and Andrew Kiss

The third session was a panel on Generative AI and Large Language Models moderated by Katherine Bode with Thang Bui, Rahul Shome, Yaya Lu, Adrian Mackenzie and Milicent Weber as panellists. This was followed by breakout sessions on the Future of HMI, focusing on AI in government and policy, active research topics on generative AI, and generation questions on AI. HMI Director, Lexing Xie, ended the workshop with summaries of the breakout sessions and imagining integrated AI.

In its initial launch in 2018, HMI’s focus was on research of ethics and the morality of AI to encourage cross-college collaboration. As AI evolved, so has HMI’s scope. Four years into the research initiative, HMI has produced work in the social impact of AI implementation, diagnoses of complex infrastructure, social media, and market attention. It has also explored how AI can contribute to the general understanding of STEM and the public good. Moving forward, HMI will broaden its scope towards integrated AI. This continues to fall in line with HMI’s mission of transdisciplinary research in addressing social and scientific problems, as well as supporting creative and cultural practices.