On Friday, November 17th, Digital Garage, a leading Technology conglomerate based in Tokyo and San Francisco, focusing on Fintech/Adtech and investing globally in early to later-stage technology startups, hosted a conference about “Building Useful and Trustworthy AI.” Several experts came together to discuss how reliable AI should be implemented in society, AI regulations, and technologies to be combined with AI that takes approaches other than large language model (LLM). Our board member Ray Hatoyama, CEO of Hatoyama Studio, also joined as a panelist.
Rocky Mitsuhashi, President of Digital Garage US, opened the event by emphasizing how Digital Garage believes that bridging technology and innovation in San Francisco and Tokyo is crucial. He highlighted that AI is a tremendous, amazing technology that can change society, but how it is used depends on us.
Joichi Ito, Co-Founder and Senior Managing Executive Officer of Digital Garage, described in his keynote address that in a few decades, AI should be thousands of times better. People are worried that AI will become too powerful and kill people. AI hallucinations, where the AI gives a confident response not backed by training data, should be fixed in a few years. He highlighted safety concerns including the fear that if AI is better at reprogramming than humans, AI could destroy humanity. He discussed other issues such as how AI can increase the risk of bioterror, terrorism involving the intentional release or dissemination of biological agents, if knowledge of how to build biological agents is available through AI. He highlighted risks with open source AI without guardrails could also create problems. For example, Generative AI can be used to generate new and complex types of malware, phishing schemes, and other cyber dangers that can avoid conventional protection measures. Such attacks may have significant repercussions like data breaches, financial losses, and reputational risks.
Finally, Mr. Ito’s comments referenced the recent discussions at G7 in San Francisco about AI governance. At G7, it was reported that the United States and China have agreed to create a working group to start a dialogue on artificial intelligence. It has been reported that the US administration is concerned about China’s use of AI, especially for military purposes. Mr. Ito commented that nobody in Japan had the job to be responsible for AI and safety, although METI and various organizations were addressing specific topics.
From the perspective of a Gen Z attendee, AI is both exciting and frightening. With exponential growth, will AI destroy humanity or make it better? AI is completely unpredictable, so no one knows where it will go from here.
Alexander Gopal
