On September 9, Liu Yuanchun, president of Shanghai University of Finance and Economics, pointed out at the 2026 Inclusion · Bund conference that there are significant differences in the economics community about how much AI can drive growth. Different studies estimate AI's contribution to total factor productivity over the next 10 years, ranging from 0.07 percentage points to 1.3 percentage points per year. The difference stems from different assumptions about AI's replaceable tasks, and there is no need to fully believe it. Regarding the current AI valuation bubble, he believes that the bubble is not scary. The key depends on whether infrastructure such as computing power and data centers formed during the bubble can be settled into real production capacity. Taking the Internet bubble as an example, asset price adjustments did not stop technological development, and the remaining physical capital still supported subsequent growth. Liu Yuanchun believes that judging whether AI investment is a bubble depends not only on valuation, but also on whether capital accumulation can cross the critical point and raise long-term output levels.

On September 9, Liu Yuanchun, president of Shanghai University of Finance and Economics, pointed out at the 2026 Inclusion · Bund conference that there are significant differences in the economics community about how much AI can drive growth. Different studies estimate AI's contribution to total factor productivity over the next 10 years, ranging from 0.07 percentage points to 1.3 percentage points per year. The difference stems from different assumptions about AI's replaceable tasks, and there is no need for full faith. Regarding the current AI valuation bubble, he believes that the bubble is not scary. The key depends on whether infrastructure such as computing power and data centers formed during the bubble can be settled into real production capacity. Taking the Internet bubble as an example, asset price adjustments did not stop technological development, and the remaining physical capital still supported subsequent growth. Liu Yuanchun believes that judging whether AI investment is a bubble depends not only on valuation, but also on whether capital accumulation can cross the critical point and raise long-term output levels.