Stanford University Tech Forum: AI Development Should Enhance Human Capabilities

The rapid development of artificial intelligence has permeated various areas of social life, with its exponential growth in applications and content sparking public concerns about ethics, data security, and regulation.

On the afternoon of May 31st last week, Stanford University hosted the 26th New Frontier in Computing (NFIC) forum, focusing on the ethical issues surrounding generative artificial intelligence.

Dr. Ricardo Baeza-Yates, the research director of the Institute for Experiential AI of Northeastern University, gave a speech addressing the limitations of generative artificial intelligence. He expressed concerns that such AI cannot learn things not present in the data. He emphasized the need for AI development to respect human dignity and incorporate accountability and responsibility, aiming to enhance human capabilities rather than replace them.

Dr. Baeza-Yates also highlighted a global ethical issue, stressing the importance of integrating ethics education from the first year of university across all disciplines to raise awareness among students about the significance of ethics, akin to how doctors understand its importance.

He stated, “I am not worried about artificial intelligence; I am worried about our intelligence. If something goes wrong, it is our fault, not AI’s fault.”

Dr. Vishnu S. Pendyala, a faculty member in the Applied Data Science Department at San Jose State University and chair of the IEEE Santa Clara Valley Section, discussed ethical concerns related to generative artificial intelligence. He pointed out the rapid development of generative AI outpacing human intelligence evolution. While AI is reaching human levels in various fields, it also brings along a growing number of issues.

He raised several key questions. Firstly, modern large language models have billions of parameters, making it impossible for humans to predict or control the behavior of AI models. Secondly, intense competition among companies leads to products being released without thorough testing.

Furthermore, he expressed concerns about the fact that most large language models are primarily trained in English, while there are approximately 7,000 languages globally. He warned that AI development could potentially bring disaster and climate change issues, stating that “artificial intelligence is a double-edged sword. If it falls into the wrong hands, it could cause massive destruction.” He believes that strong regulation is crucial in the path of AI development, noting the current insufficiency of government oversight.

During the forum, Dr. Jesmin Jahan Tithi from Intel discussed using Z-Inspection® to ensure the credibility of generative artificial intelligence. Dr. Sindhu Joseph, founder and CEO of CogniCor, talked about the impact of generative AI on the workforce. Dr. Ruksana Azhu Valappil emphasized the importance of implementing ethical and responsible AI in the healthcare sector. Dr. Yang Liu explored the challenges and progress in establishing trustworthy large language models. Additionally, Dr. Pratik Desai shared insights on utilizing AI to facilitate climate-resilient agricultural adaptations.

The forum was co-organized by the North America Taiwanese Engineering & Science Association (NATEA) and the IEEE Santa Clara Valley Section, with IEEE serving as the rotating chair of the 2024 NFIC.

Attendees included NATEA President Yang Yaohong, Secretary General Chang Zhiyang, Silicon Valley Chapter President Lee Xianghe, Silicon Valley Chapter Vice President Lin Youli, former Presidents and founders Yu Jinbang and Chen Yanguang, as well as Zhou Xinjie, Ye Junxiong, and Ni Wenjun. Additionally, Taiwanese IEEE Sixth District Chair Joseph Wei and Overseas Chinese Affairs Commissioner Lin Zhengyuan were also invited to attend.