At the 2026 Yabuli Forum Summer Annual Meeting, held in Chengdu, Sichuan from September 4th to 6th under the theme "Enterprise Innovation and Cycle Crossing," Zhou Qi, Managing Director of Jinqiu Fund, shared his perspectives on AI industry investment, outlining a four-dimensional framework for evaluating potential projects.

According to Zhou Qi, the primary consideration is the selection of the赛道 direction, which serves as the foundational question. He noted that with the continuous upgrading of computing power, many deep learning algorithms that were previously impractical to run can now be deployed. Currently, large models are evolving along the scaling law, relying on the accumulation of computational power, parameters, and data for iteration. The fund prefers to invest in tracks that have the potential to define the industry landscape over the next three to five years.

Once the direction is determined, the second critical factor for evaluation is the founding team. Zhou Qi observed that the investment process has led to a classification of two types of founders who are considered promising. The first category comprises industry veterans with deep, long-term experience, particularly those who have been entrenched in sectors such as semiconductors and have now made a determined commitment to entrepreneurship. Simultaneously, the fund has invested in a substantial number of young entrepreneurs, evaluating them primarily on two capabilities. The first is an exceptional ability to learn quickly. The second is resource acquisition and growth potential. For a young founder, the question is whether the resources they can mobilize, their growth trajectory, and their judgment at this stage of their career significantly surpass those of their peers. If so, this indicates that the team possesses considerable development potential.

The third dimension involves the internal investment conviction. Zhou Qi explained that investment decisions are inseparable from information assessment, ultimately relying on the team's ability to form a firm and confident judgment. The emergence of AI tools has lowered the research threshold in frontier fields. New and emerging tracks often feature multiple parallel technological routes. At this point, a choice must be made: to bet on a single technological pathway or to construct a portfolio covering multiple routes, wagering that at least one company in the track's endgame will emerge victorious. This, he said, is the process of building an investment conviction.

The fourth point is project valuation. Zhou Qi pointed out that currently, many early-stage enterprises are initiating two or even three consecutive funding rounds before their business has even launched. Regarding this phenomenon, he believes that moderate froth is a positive signal, as it indicates that the industry has gained the confidence of capital. He drew a parallel to the wave of domestic semiconductor substitution that began around 2018-2019. At that time, the industry's prospects were not entirely clear, yet capital's willingness to bet on early-stage projects provided a significant boost to industrial growth.