At the 2026 Sohu.com Technology Annual Forum, Sohu.com founder, Chairman, and CEO Zhang Chao Yang delivered a presentation on physics knowledge. Discussing the impact of AI on science communication, Zhang stated that while AI technology offers great convenience for popularizing basic science, it requires proper guidance for the audience.

Using his own physics lectures as an example, he noted that when researching topics like satellite synchronous orbits or lunar free fall, AI tools can quickly provide formulas and directional guidance, compressing what was once days of data compilation into a single day.

"High-quality science communication should not stop at information transfer; it must also guide the audience to calculate and research personally," Zhang said. "Much popular science content caters to curiosity by using vague analogies instead of explaining principles, which actually deviates from the essence of science communication."

In his view, science communication in the AI era should place greater emphasis on logical deduction and interactive experience. Through a model of "hardcore derivation plus live interaction," the audience can truly grasp the principles of knowledge, moving beyond superficial "wow" moments of understanding.

Addressing the impact of excessive AI tool use on human thinking, Zhang cited a research case from the MIT Media Lab, indicating that over-reliance on AI can indeed lead to a decline in proactive thinking abilities. Experiments showed that individuals who completed writing tasks independently without any tools exhibited the highest level of brain activity, while testers using AI showed only half that level of activity.

Zhang emphasized that while AI can efficiently integrate information, humans must maintain their capacity for original thought and hands-on ability. "For instance, when I prepare my physics lectures, AI can quickly provide formulas, but only through personal derivation and calculation can one truly understand the knowledge." He called for maintaining a "proportion of thinking" in the AI era to avoid the degradation of brain function from passively receiving information, asserting that active exploration and deep thinking are irreplaceable core human competencies.