In China, differing views on AI from the founder of a LLM company and an investor have sparked a major discussion

Recently, Tencent News interviewed articles with Yang Zhilin and Zhu Xiaohu, sparking discussions on the direction of AI.

Yang Zhilin is the founder of Moonshot AI, which has just raised one billion US dollars.

Yang Zhilin's report: 月之暗面杨植麟复盘大模型创业这一年：向延绵而未知的雪山前进

Zhu Xiaohu, on the other hand, is an investment partner at GSR Ventures, and they also have very impressive investment performance

Zhu Xiaohu’s report：朱啸虎讲了一个中国现实主义AIGC故事

The two of them have their own different views on many issues; below are their differing opinions.

Please continue to look at the translated English version.

The text is a comparison of views between Yang Zhilin and Zhu Xiaohu on different matters:

1. On the view of large model entrepreneurship:

- Yang Zhilin: Firmly optimistic about large model entrepreneurship, believes it is the only meaningful thing in the next decade, and sees it as a system combining science, engineering, and business, requiring long-term investment and persistence, with the ultimate goal of achieving AGI (Artificial General Intelligence).

- Zhu Xiaohu: Not optimistic about large model entrepreneurship, believes that current large model companies lack scenarios and data, are overvalued, and face competition from open-source models, making investment risky. He prefers to invest in application layer projects that can be commercialized and monetized quickly.

2. On the view of open-source models:

- Yang Zhilin: Believes open-source models are behind closed-source models, and this gap will continue because the open-source development method and the concentration of talent and capital cannot match those of closed-source models.

- Zhu Xiaohu: Believes open-source models will gradually catch up with closed-source models because the technology iteration curve will slow down, and the open-source community has more developers involved. He thinks Chinese developers prefer open-source models because they don't have to worry about being copied.

3. On the view of AGI:

- Yang Zhilin: Firmly believes AGI is the future and makes it the long-term goal for company development. He thinks AGI will change the world and hopes to achieve this goal through technological breakthroughs and the combination of user products.

- Zhu Xiaohu: Skeptical about AGI, believes it cannot be achieved at least within 5 to 10 years. He is more concerned with applications that can be commercialized in the short term and thinks investing heavily in AGI research at this stage is too risky.

4. On the view of the Chinese and American large model industry:

- Yang Zhilin: Believes there won't be much difference between China and the US in terms of basic general capabilities, but the differentiation of applications based on general capabilities is more likely to occur. He also believes that Chinese companies have the opportunity to take the lead in certain areas.

- Zhu Xiaohu: Believes there is a significant gap in the large model field between China and the US, with the US leading in underlying technology, while China has more advantages in application scenarios and data. He suggests that Chinese startups...

5. On the view of investment strategy:

- Yang Zhilin: Pursues long-termism, willing to invest a lot of money and resources in frontier exploration, and believes that technological breakthroughs will ultimately bring commercial returns.

- Zhu Xiaohu: Seeks short-term returns, prefers to invest in projects that can be commercialized and monetized quickly, and emphasizes that startups need to control costs and not burn money.

6. On the view of Sora:

- Yang Zhilin: Believes Sora is an important milestone that represents a huge advance in video generation technology, and thinks it can be used to enhance the understanding of multimodal inputs, as well as to bridge the digital and physical worlds.

- Zhu Xiaohu: The article does not explicitly state his view on Sora, but he is optimistic about the prospects of video generation technology and believes that Chinese companies have the opportunity to lead in this field.

7. On the view of the relationship between startups and giants:

- Yang Zhilin: Thinks that giants and startups have different goals and strategies, but there is a relationship of competition and cooperation between them.

- Zhu Xiaohu: Believes that large model entrepreneurship companies may eventually be acquired by giants, but the acquisition price will not be high because of the severe homogenization of large model technology.

8. On the view of the risks and challenges of entrepreneurship:

- Yang Zhilin: Accepts the risks and challenges of entrepreneurship calmly, and says he will charge forward fearlessly.

- Zhu Xiaohu: Thinks that under the current macro environment, startups need to focus more on controlling costs and self-sustaining abilities to survive in the market competition.

9. On the outlook for the future:

- Yang Zhilin: Optimistic about the future, believes AGI can push human civilization into the next stage.

- Zhu Xiaohu: Cautious about the future, thinks technological development has uncertainties, and startups need to be more realistic and pragmatic.

10. On the view of talent:

- Yang Zhilin: Values talent density and believes that the company's ceiling is determined by the ceiling of its people. He first looks for geniuses, and then supplements other dimensions of talent to build a diverse team.

- Zhu Xiaohu: Thinks that the talent in the large model field is severely homogenized, and startups need to find talent with business acumen and management and sales capabilities to win in market competition.

11. On the view of commercialization:

- Yang Zhilin: Believes that commercialization is the means and goal of achieving AGI, but should not give up long-term goals for the pursuit of short-term commercial interests.

- Zhu Xiaohu: Thinks commercialization is the only standard to test the value of large model companies and emphasizes that startups need to find quickly monetizable scenarios and applications.

Summary: Yang Zhilin and Zhu Xiaohu represent two different investment philosophies and judgments on the development path of AI. Both views have their rationality, and time will tell which judgment is more accurate.