Sunday, May 26, 2024

The future of generative AI in Korea

 During the past several semesters, including Spring 2004, I've been teaching an undergraduate course entitled Modern Digital Technologies and Innovation (EST 204 at Stonybrook University).  Not surprisingly, the public introduction of generative AI with the release of chatGPT in the Fall of 2022 had a big impact on my approach to this class.  I've been encouraging students to explore the possibilities of generative AI for both good and bad.  The "emergent properties" sometimes generated by AI models may sometimes be good, but may also on occasion be hallucinations that may have various adverse impacts. 

The rapid spread of generative AI clearly suggests that there is an urgent need to train large numbers of students in what we may call generative AI literacy.  In my class, students work in small teams of 5-6 students each.  They are asked to select one of the sustainable development goals and to use generative AI and the design thinking process to develop solutions to help achieve the goal. 

The massive electric power requirements to implement generative AI on a large scale using silicon chip based computing and large data centers suggests that we are only at the beginning of a process that may eventually revolutionize the role of technology in sustainable development.  The human brain, although it runs on significantly less power, has far more capabilities than even the largest and latest AI models.

South Korea harnessed the digital network revolution for national development faster and more effectively than any other country in the world.  However, it did so with relatively more emphasis on hardware and exports than on software and services.  The arrival of the generative AI era suggests that future developments will depend largely on AI generated software and services across all sectors of society, industry and government.  In future posts I will try to break down some of the issues and possibilities.