Author: Techub News Compilation

Introduction

At the OpenAI DevDay 2025 event, CEO Sam Altman gave an exclusive interview. This is Altman's first in-depth public discussion following the release of a series of major updates from OpenAI, including the ChatGPT app store, the powerful Agent Builder tool, and the public testing launch of the video generation model Sora. As one of the most influential leaders in the AI field today, Altman not only interpreted the thinking behind the latest products during the conversation but also shared his profound insights on AI technology development trends, the definition of AGI (Artificial General Intelligence), the social impact of AI, and future forms of work. This conversation provides valuable first-hand materials for us to understand OpenAI's strategic direction and the future of the entire industry.

Summary

• Agent Builder is sparking the "no-code revolution" in AI, significantly lowering the threshold for building intelligent agents and leading to a surge in global software output.
• Sora, as an important step towards AGI, is being explored by OpenAI for its business model, believing that video will become a crucial interactive interface.
• The AGI definition that Altman values most is "the ability to make brand new scientific discoveries, expanding the boundaries of human knowledge."
• “Zero-person companies” (companies completely operated by AI agents) may emerge in the future, but this will still take several years.

Agent Builder: AI's "No-Code Revolution" and the Prospects of Zero-Person Companies

Throughout the interview, Altman repeatedly mentioned OpenAI's latest Agent Builder, viewing it as a significant paradigm shift. He recalled the introduction of GPT Builder at the first DevDay two years ago and pointed out that the greatest advancements come from leaps in model capabilities. Now, with Agent Builder, even ordinary "knowledge workers" can build and deploy quite complex intelligent agents in just a few minutes, with almost no coding necessary, using visual interfaces, uploading files, and connecting data sources.

“It feels like a tectonic plate movement,” Altman remarked, “I haven’t fully grasped what this means yet.” He predicts that Agent Builder will significantly accelerate software production efficiency, the global output of software will sharply increase, and the time required for testing and optimizing ideas will continuously shorten. This is not just a tool, but the beginning of a new productivity revolution.

When asked whether this enhancement of capabilities would make a “zero-person company” possible, Altman offered a cautious yet optimistic outlook. He revealed that he and his friends used to bet on when a “one-person billion-dollar company” would appear, and now they have started informally discussing the arrival of “zero-person companies.” He expects this will still take “several years” but considers it remarkable that “we can now seriously discuss the arrival of this day” in itself.

Regarding the current development level of AI agents, Altman believes that we are “not far” from agents that can truly work independently for a week without human feedback. He noted that not long ago, people were amazed that AI could complete tasks that lasted a day, and now, agents capable of handling week-long tasks seem to be just around the corner. The key technological bottlenecks to achieving this goal are: smarter models, longer context processing capabilities, and better memory mechanisms.

Sora, AGI, and the Social Integration of AI

As one of the most talked-about AI products recently, Sora's public testing sparked a lot of discussions and memes, including many generative videos featuring Altman himself. In response to this, Altman showed an astonishingly open attitude. He admitted that he initially thought it would be a very strange experience, but after browsing through hundreds of Sora videos about himself on social networks, that “strange feeling” disappeared after just a few minutes. He compared this to the first time he rode in a Waymo self-driving car—“the weirdness was temporary.”

However, Altman also expressed concerns about the phenomenon of Sora’s watermark being removed. He explained that OpenAI chose to release this type of technology now precisely because they foresee that in the coming months or years, anyone could use open-source models to create untraceable, watermark-free deepfakes. By releasing products with protective measures early, society and new technologies can evolve together, providing a form of “vaccination” in advance to adapt to this inevitable trend.

Regarding the relationship between Sora and AGI, Altman clearly stated that exceptional video generation capabilities are an important step on the road to AGI. This not only concerns the video content itself but also the spatial reasoning abilities and world model construction that are crucial for achieving true AGI. He envisions a future where real-time video stream interactions with AI will create entirely new user experiences. However, Sora is still a product that has only been live for a few days, and its business model (such as charging per generation, potential ad models, or revenue sharing with creators) is still being explored.

When discussing the definition of AGI, Altman stated that although there are various definitions, what he is personally most concerned about is whether AI can make “brand new scientific discoveries” that expand the total knowledge of humanity. He mentioned that we are already seeing early signs of AI proposing new methods or making small discoveries in certain scientific fields. This gives him a sense of “great significance.” He predicts that society will quickly adapt to the fact that AI makes scientific discoveries, just as people were only surprised for two weeks after GPT-4 easily passed the “Turing Test” in public perception before they started demanding it to be “faster and better.”

The Impact of AI on Work, Building Advantages, and Regulatory Thoughts

Regarding AI’s impact on work, Altman adopted a historical perspective. He used the example of farmers to point out that people in the past might not have considered many of today’s knowledge workers’ tasks as “real labor.” Similarly, new jobs that emerge in the future AI era may not “seem like work” to us today. He acknowledged that there are short-term concerns during the transition period when AI rapidly replaces certain knowledge work, but in the long run, he believes that human drive will lead us to discover a multitude of new and meaningful things to do, and the “contract” of society will change accordingly.

Regarding how entrepreneurs can establish lasting competitive advantages in the AI era, Altman believes there is no one-size-fits-all answer. Traditional advantages such as network effects, brand, data, and market position remain effective, but the best advantages are often unique paths discovered by founders concerning their own products, technology, and market timing. He quoted a favorite business maxim: “Let tactics evolve into strategy.” — that is, starting from effective specific actions often leads to the natural formation of a winning strategy. He cited the “memory” function of ChatGPT as an example; this now crucial competitive advantage wasn't initially planned at launch.

When asked what global policy he would implement if he could, Altman’s answer was: to establish a global safety framework for cutting-edge AI models to reduce catastrophic risks. He emphasized the importance of establishing good testing and evaluation frameworks, which could serve as a starting point for global cooperation.

Product Philosophy and Future Interaction: ChatGPT is not a "Universal App"

Regarding speculation about whether OpenAI is building an "everything app" similar to WeChat, Altman clearly denied. He stated that the market environment in the United States is different, and OpenAI's goal is to create a "super AI assistant", rather than an all-encompassing platform. This is why Sora was chosen to be released as a standalone application instead of being directly integrated into ChatGPT. He believes that users have different mental models for "personalized AI assistants" and "entertainment applications," and forcing them to merge would create an incoherent experience.

When discussing future interaction methods, Altman believes that voice will be an important but not the only interface. In suitable scenarios, voice interaction feels very natural. He mentioned that smart speakers are often joked about, but many people still love and use them, while the current AI and infrastructure are not yet sufficient to make such devices truly outstanding. He hinted that OpenAI is exploring a new device form aimed at achieving the ideal user experience for interacting with computers through natural means like voice, but this requires time and patience to refine.

Finally, when asked what he would build if he could return to age twenty with today’s knowledge, Altman expressed envy towards today’s young entrepreneurs. He believes the opportunity space now is “incredibly vast.” Although his own work at OpenAI occupies all his energy, he feels very excited about what many developers present are building.

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。