Not a Consensus of Consensus.

Author: Jingyu

The Shanghai World Artificial Intelligence Conference (WAIC) has reignited!

On July 26, 2025, this year's WAIC attracted over 1,200 guests from more than 30 countries and regions, including 12 Turing Award and Nobel Prize winners, over 80 academicians from China and abroad, and representatives from several top international laboratories.

During the opening event on July 26, 2024 Nobel Prize winner Geoffrey Hinton, regarded as a master in the AI field, delivered a speech. This well-known scholar, who has consistently advocated for the "AI threat theory," reiterated the dangers of unregulated AI development and called for the establishment of a "safety network" for AI research globally.

Additionally, Yan Junjie, founder and CEO of the rapidly rising domestic AI startup MiniMax, stated in his speech that the growth of increasingly powerful AI has almost no limits, and with the decreasing costs of training, "future AI will be more inclusive."

Meanwhile, Peng Zhihui (Zhihui Jun), co-founder and CTO of the domestic robotics startup ZhiYuan Robot, which has just been rumored to be "going public," showcased their robot product Lingxi X2 on stage, performing a segment of "crosstalk," providing a more tangible representation of the "partnership" between robots and humans.

Lions or Babies, Both Need "Supervision"

Even though disciples occupy half of the Silicon Valley AI tech circle, Geoffrey Hinton, the "master" of the AI field, has consistently sung a different tune from his industry peers, firmly maintaining the "AI threat theory."

At the WAIC conference on July 26, Hinton once again expressed his concerns about the rapid development of AI in his speech.

After briefly describing the development of AI technology over the past 30 years to the current large model stage, Hinton believes that the way large models understand language is similar to that of humans.

"Humans could be large language models and may also generate hallucinations, creating many illusory languages," Hinton's insight aligns closely with the so-called "human-machine" memes prevalent on social media today.

However, the problem is that compared to the "carbon-based brain" of humans, the "silicon-based brain" equipped in AI has inherent advantages in storage, replication, and "instant transmission." This suggests that with technological advancement, it is widely believed that the emergence of AI that is smarter than humans is merely a matter of time. As a form of existence, these AIs, representing "intelligent agents," will undoubtedly demand "survival" and "control."

Hinton likens current AI to a "three-year-old child," easily manipulated by humans, but the future is uncertain. He also compares current AI to a young lion, stating that there are only two possibilities for raising a lion: "either train it not to attack you or eliminate it."

From the current global AI progress perspective, no country can truly "eliminate AI" by halting technological development. The only path left is for the world to establish an AI safety agency to train AI to be "benevolent."

"How to train an AI that does not want to dominate humanity is the ultimate question facing humanity," Hinton concluded.

Some Seek Regulation, Others Advocate for "Deregulation"

Such remarks from a master are thought-provoking at WAIC, but in North America, where Hinton is based, the situation seems somewhat "short-sighted." Companies like OpenAI and Anthropic, which are associated with Hinton's students, are already valued at hundreds of billions of dollars, not to mention the significant investments Silicon Valley venture capitalists have made in AI startups over the past two years.

A prominent manifestation of this is that as AI companies increase their lobbying expenses in the U.S. political arena, American regulators have officially loosened their control over AI development.

Also this week, on July 23, local time, then-President Donald Trump released the AI Action Plan. In this document, U.S. regulators outlined measures to ensure that AI maintains its leading advantage in the U.S. from aspects such as data, standards, and talent:

1. Increase R&D Investment: Significantly boost long-term federal investment in foundational and applied AI research, especially in next-generation AI, AI safety, and trustworthy AI.

2. Unleash AI Data Resources: Promote the safe opening of vast datasets owned by the federal government to AI researchers and the public, providing high-quality "fuel" for model training.

3. Set AI Technical Standards: Led by the government, collaborate with industry and academia to establish benchmarks, standards, and norms for global AI technology, ensuring AI systems are safe, reliable, explainable, and fair.

4. Cultivate an AI-Ready Workforce: Reform STEM education, promote apprenticeship and retraining programs, and attract and retain top global AI talent to build human resources for the AI economy.

5. Strengthen International Collaboration: Establish AI alliances with allies and partner countries to jointly formulate rules, counter the misuse of AI by "authoritarian states," and promote open and democratic AI applications.

6. Protect Key Technologies: Enhance protection of critical U.S. AI technologies, algorithms, and hardware (especially semiconductors) through export controls and investment reviews to prevent technology from flowing to strategic competitors.

It is evident that the U.S. is giving the green light to the development of AI technology while simultaneously using geopolitical issues to prevent competitors from being on the same starting line.

The "Experience Era" of Robotics Has Arrived

At this year's WAIC conference, humanoid robots were undoubtedly the most captivating highlight, without exception.

During the main forum, Peng Zhihui, co-founder of ZhiYuan Robot, presented the company's product Lingxi X2 on stage, performing a segment of "human-machine crosstalk."

Crosstalk emphasizes speaking, learning, joking, and singing, while the interaction between robots and humans emphasizes "long live understanding." Lingxi X2 stated on stage that collaboration between humans and robots must be based on "consensus." But how to establish consensus between robots and humans and break through the key to human-robot collaboration? Zhihui Jun mentioned that this is the track the company aims to delve into, emphasizing the importance of walking this path with more peers.

Thus, Zhihui Jun also announced the launch of the "ZhiYuan Lingqu OS" open-source plan on-site, hoping to promote the integration of the current robotics system ecosystem and breakthroughs in embodied intelligence new technologies together with more people.

Seemingly to endorse embodied intelligence, Richard Sutton, a professor in the Department of Computer Science at the University of Alberta and 2024 Turing Award winner, participated in the conference via video link. He believes that the data currently used to train large models is nearly exhausted. However, there is no need to be disheartened, as this signifies that the next era of AI—the Era of Experience—is about to arrive.

Unlike the past, where AI was trained using "static" data, the next phase will allow AI to gain knowledge and enhance capabilities by "experiencing" the external environment and objects, much like human infants. Although this goal is still somewhat distant, many robotics startups are indeed continuously training and learning in the "physical world."

This is also why top scholars in the industry, including Fei-Fei Li, have transitioned from "AI" to "Physical AI," emphasizing that for artificial intelligence to truly enter the real world, it must understand and learn the entire world from a three-dimensional perspective.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。