Activity
IEK Symposium on Trends of AI
A panel discussion was held in IEK’s AI symposium.
Artificial intelligence (AI) has been seen as a revolutionary technology to transform industries and bring new market opportunities. The Industrial Economics and Knowledge Center (IEK) of ITRI hosted a symposium on Trends of AI Technologies and Applications in Taipei on June 26, inviting AI experts from Stanford University and Silicon Valley to share their experiences on AI research and venture development. The event attracted over 300 participants from industry and academia.
ITRI established a collaboration platform with Stanford University in 2004, which has led to annual visits and conferences that have enhanced the link between Taiwan and Silicon Valley. IEK General Director Stephen Su remarked that according to IEK observations, the AI industry in 2017 has entered a hastened phase of growth in areas such as image recognition, speech recognition, and medical diagnosis, triggering a reorganization of the industrial chain. Consequently, he suggested that Taiwan should convert “AI x” into an “x factor,” in which the integration of software and hardware drives innovation. Taiwan’s industrial expertise and ICT technology should be integrated with AI into various applied service systems. Examples include “AI x Smart Transportation,” “AI x Smart Retailing,” “AI x Smart Manufacturing,” “AI x Smart Health,” and “AI x Smart Travel.”
Prof. Chuck Eesley from Stanford University gave a speech on Venture Investment in AI Startups and Future Opportunities, covering the economics of AI, top venture capitals’ investments in AI/machine learning startups, and recent IPOs/acquisitions. He also emphasized some future commercial opportunities for AI developments, and highlighted the Asia Pacific region as having the biggest potential.
The presentation by Dr. Nicolas Pinto, Founder of AI startup Perceptio, explored the intertwined relationship between reverse and forward engineering of intelligence in the context of mobile devices/hardware. He pointed out that mobile data, which provides a better snapshot of real life, may be more appropriate for building future AI machines. He further indicated that the next milestone of AI is to develop computational vision neuroscience for image recognition by learning from human brains. Dr. Pinto hoped that his study can close the gap and manage the expectations set by an over-promising AI hype that is detaching from reality at an alarming rate.
Prof. Silvio Savarese, Director of SAIL-Toyota Center for AI Research of Stanford University, said that despite the tremendous progress that machine learning, computational vision and perception robotics have made over the past decade, AI is still far from exhibiting the level of visual intelligence and behavioral traits—common sense reasoning, decision-making, adaptation and generalization, to cite a few—that humans are capable of as they observe, navigate and interact with the real world. He discussed challenges and opportunities in modern AI and showed its potential impact on applications related to autonomous driving, assistive robotics, robotic mobility, smart environments, augmented and virtual reality, as well as large-scale information management in the construction business.
“We need to think about humans as well—how we can work with and have a better understanding about machines,” said Prof. Jane Yung-Jen Hsu from National Taiwan University. She suggested that we need to develop intelligent technologies that can enhance human-machine collaboration and real-time interaction, which can be used to assist in areas such as smart manufacturing. She further commented that researchers in Taiwan can collaborate with industry to provide vertical integration solutions for AI.
The symposium also invited Dr. Chih-Han Yu, Founder and CEO of Appier, to share his experience as both an AI scientist in the lab and an entrepreneur in the actual market. Appier was selected by Fortune magazine as one of the 50 companies leading the AI revolution. Dr. Yu talked about how he has leveraged the innovative capacity and talents in Taiwan and led his team to develop the world’s leading AI applications and services. “After 2020, I think AI will not be merely a category; it simply exists everywhere. Companies will take it as a core competency to drive business efficiency or as their business consultant to improve the decision-making process,” he said.