Taipei, Sept. 4 (CNA) Semiconductor industry leaders discussed whether the artificial intelligence (AI) boom is worth the investment and who will benefit most from it, in a rare exchange Wednesday at an annual semiconductor industry expo in Taipei.

In the AI Chip Fireside Chat — called “the chat of the century” in Chinese — and part of the CEO Summit at SEMICON Taiwan, ASE CEO Tien Wu (吳田玉), the moderator, posed hard questions to top executives from Google, Taiwan Semiconductor Manufacturing Co. (TSMC), and Samsung Electronics.

Y.J. Mii (米玉傑), executive vice president and co-chief operating officer of TSMC, Hamidou Dia, vice president of Applied AI Engineering at Google, and Lee Jung-bae, Samsung Electronics corporate president, were the three participants in this rare exchange.

Leading an IC packaging and device manufacturing company himself, Wu said while the “insatiable demand” of chips brought by the AI boom “makes us happy, it also makes us sleepless.”

As corporate executives estimate the trillions of dollars AI could contribute to the global economy by 2030 or 2045, Wu said he wanted to ask the industry leaders whether they believe the so-called AI boom is really worth the investment from the hardware side.

“Is the AI [boom] really that big and worth the pain?” he asked.

In response, Dia said compared to the internet and social media platform Facebook which took seven and three years, respectively, to reach 15 million users, generative AI chatbot did the same “in five weeks.”

This showed that it was “the biggest technology shift, the biggest platform shift we’ve seen in our lifetime, and it is also the fastest moving technology we’ve seen,” Dia added.

Lee from Samsung agreed, calling himself a “cautious optimist” and saying that the forecasts made by experts for the internet and smartphone booms “were failures” as in the end the economic progress made by the two innovations “were higher than predictions.”

Mii concurred, saying that “AI is a huge opportunity for this industry — the ChatGPT we currently see is only one type of AI,” and it can be applied in many aspects of life.

Wu also asked whether all economies, regardless of size, will benefit from the AI rise.

Mii said he believes everybody will benefit. “It might go to the largest economy first, such as the United States, with all of its applications and software [capabilities], but it will spread out,” he added.

Lee said different countries may benefit at different levels and times.

The AI expert from Google took Africa, where he is originally from, as an example, saying that chatbot has been widely used on that continent.

“For the first time, we are providing models that are pre-trained on all publicly available information around the world,” he said.

“That’s a big deal,” as it means the entry barrier to the AI world is already lowered by this pre-training, Dia said.

Benefits for the whole also come from the fact that AI will help advance medicine and accelerate drug discovery, he added.

However, Wu countered with the idea that as the U.S. controls the majority of data, models, and design capabilities, “won’t it be the most likely country to profit from the efficiency jump and widen the disparity gap?”

Mii said the U.S. and China with the greatest data and the largest markets are likely to lead, while Taiwan and South Korea are good at certain parts and might benefit partly.

Dia differed. Models are all pre-trained with publicly available information, and after that, all countries can use their own data to produce specialized applications, he said.

“Like Taiwan has more data about manufacturing than anybody else in the world” and will be able to make specialized models, Dia said.