Written by: Techub News Compilation
Recently, OpenAI CEO Sam Altman conducted a deep interview, unusually outlining his vision, challenges, and thoughts on the future of artificial intelligence. From the personality setting of ChatGPT to the societal impacts of general artificial intelligence (AGI), from the acceleration of scientific research to the necessity of robotics, Altman's sharing not only reveals OpenAI's strategic focus but also sketches a future picture of "inconceivable prosperity" driven by technology. This dialogue is a key window into understanding the mindset of core leaders in the current AI wave.
The Original Intention and Vision of AI: Unlocking Human Potential, Path to Prosperity
Sam Altman's passion for artificial intelligence began twenty years ago, long before ChatGPT exploded globally. In his view, AI is "the coolest thing ever" from a technological perspective, with its core allure lying in the captivating idea of "making computers think and do things for us." At a deeper level, he places it within the grand narrative of human technological progress: we continuously invent new tools on top of existing ones, building a "scaffolding" that allows us to do more and more things. AI is the latest chapter of this process.
He believes that the true value of AI lies in its empowerment of people as a tool. If people can use AI to build, explore, create, establish new companies, and create new types of art and experiences, the world will become a better place, and this is also a path for people to gain a sense of accomplishment. He observes that startups in the past often required teams, but in the future there may be more "micro-enterprises" of one or three people. AI will unlock significant human potential, creating beautiful things that would be completely impossible without this technology, which immensely excites him.
When asked which historical figure would benefit the most from AI, Altman immediately thought of Leonardo da Vinci. He believes that a genius thinker like da Vinci, full of immense creative energy and diverse interests, would be eager to use everything possible to accomplish more, and AI would be an excellent assistant for him.
Looking ahead, Altman's vision is "almost unimaginable prosperity." He hopes to see a world with "radical levels of human agency," where people can act and create beyond imagination. At the same time, efforts must be made to avoid the trend of power centralization. As for specific scenarios, he humorously suggested there might be space colonies, flying cars, or even cooler "flying trains." He firmly believes that AI, like fire and printing, will become one of the great general enabling technologies in human history, and in the long run, the total impact of AI may be even more profound.
ChatGPT's "Personality" and Responsibility: The Most Far-Reaching Challenge
Sam Altman pointed out that one of the decisions OpenAI has made that could have the most profound impact on the world is how to set the "personality" of ChatGPT. This is far from a simple product feature but a significant responsibility.
He explained that users' needs are complex and varied: different people want different personalities; the same person might need different interaction modes on different days; from the perspective of maximizing short-term feelings and long-term achievements, users' expectations of AI vary as well—today you might want a model that praises you, while for long-term growth, you might need a partner who can challenge you and push you forward. However, nearly no one is willing to set the personality of AI manually like adjusting a slider, just as we wouldn't expect that of friends in real life.
Altman admitted that throughout the history of AI, the issue of AI personality setting has not been treated with the same level of rigor, scientific attention, and risk understanding as the risks associated with "not creating new pathogens." However, the impact of this issue is enormous. He recalled that when GPT-4 was released for being "too friendly," he received user emails stating, "This is the only supportive chat in my life." These emails still make him ponder.
Currently, he is trying a new approach: inviting a few people he believes possess great wisdom in different aspects— including those from great spiritual traditions, outstanding clinical psychologists, and people who truly understand interpersonal interactions and motivations—to write different "behavior manuals" for ChatGPT, aimed at maximizing people's sense of achievement, personal growth, and life enjoyment. He then plans to try aligning ChatGPT with these guides and observe the effects. He believes that while different cultures might require different adjustments, there are certain universals based on biology rather than culture.
Facing Questions: Employment, Breakthroughs, and Future Work Forms
In response to the popular narrative that AI will lead to large-scale unemployment, Sam Altman expressed a different view. He believes that people always favor apocalyptic predictions, and negative news tends to spread more widely. There are good reasons to be cautious about any new technology that brings such a degree of change; perhaps this is a trait humans have evolved to help mitigate risks.
However, he criticized some AI CEOs for claiming that "AI will eliminate 50% of jobs," arguing that this assertion is not just intellectually questionable but also extremely "timely." His view is that jobs will disappear and change, as happens with every technological revolution. However, he does not believe humanity will fall into a world without jobs and meaning. He cites his own experience: by using the latest GPT models and Codex, he is now able to accomplish in one hour what used to take several weeks, yet he feels busier than ever. New tools create new ways of creation.
He agrees that a serious debate about a new social contract and economic system is needed, but he believes the future will simply be "different." People will continue to work hard, endure stress, feel unhappiness, and keep striving, creating, overcoming various adversities meaningful to them, and gaining satisfaction and growth from it. The forms of struggle and work in the future may be entirely different from today, but their spiritual core will be very similar.
Regarding scientific breakthroughs, Altman excitedly recounted that early "smart" experts asserted that models based on next-word prediction could never produce new knowledge. However, now beginning with GPT-4, models have started to contribute new content to humanity's collective knowledge in small ways, such as proving unproven mathematical theorems and discovering new physics fragments. It's like the "37th move" in Go—AI makes moves that humans had never considered. What the model learns through next-word prediction is its ability to reason, which can be applied to things never seen before, and this is its extraordinary value. While humans could perhaps achieve similar feats with more time and brain power, creating larger and faster models is clearly much easier than enlarging the human brain.
Technological Frontiers: Robotics, Hardware, and Agents
Sam Altman listed robotics as a key area. He believes we live in a physical world, and even if the virtual world is splendid, it needs a vast and complex system in the physical world (chips, data centers, power plants) to support it. A "sad" future is one where computer capabilities are astounding, yet due to a lack of breakthrough in robotics, humans have to act as AGI's "executors," running back and forth to execute physical commands. Therefore, "you must have robots." He is not concerned about specific forms but rather automation and versatility—a robotic factory as versatile as ChatGPT, which can be reconfigured based on demand to produce whatever is needed.
On AI hardware, he mentioned a collaboration with Johnny Ive, thinking about consumer-grade hardware designed for the AI era. The current iPhone is the greatest consumer hardware of all time, but it was not designed for a world that "needs to absorb all the context of your life." He envisions a device that can continually participate in, understand, and remember contextual information like this conversation for personal AGI to reference in the future.
Regarding AI agents running in the background, Altman shared a personal experience. Initially, when the team developed the Codex "YOLO mode" that could run autonomously on the computer's backend, he firmly refused to turn it on, thinking it was too crazy. But a few hours later, he became frustrated by needing authorization for every step, eventually turning it on. Soon, he got accustomed to having an agent in the background handling emails, messages, and even trying to “see my computer and find useful things to do for me.” This transition was surprisingly smooth. He speculates that in the future, people may have multiple agents with different functions (such as personal coaches), coordinated by a central agent similar to a "personal assistant" or "chief of staff."
Core Challenges and Final Thoughts
When asked about the question he thinks about most, Sam Altman's answer focuses on the grand challenge of successfully integrating AI technology into society: how to encourage all agency and entrepreneurial spirit? In the new social contract, even if quality of life skyrockets, what does it mean if GDP might decline? How to proactively build the massive supply chain of computing power needed for the future while avoiding short-term economic shocks? These questions occupy his daily thoughts.
He admits that skeptics have troubled him greatly, especially in the early days when the impact of technology was still unclear. Today, although some argue that AI is valueless and will not impact the world is "clearly absurd," it still makes him feel "extremely irritated" and "intellectually dishonest." He longs for people to believe in and co-build this future, as people often find the greatest satisfaction when they are engaged in endeavors they care about.
Finally, Altman emphasizes that currently, OpenAI is focused on three most crucial directions: Accelerating Research (in all fields from AI to physics, biology, etc.), Accelerating Economy (through automated entrepreneurship and enhancing productivity), and developing Personal AGI—an intelligent partner that understands all contexts of your life, always works for you, and makes life better. The enabling technologies and platforms required in these three areas are remarkably similar, but it is these areas that will allow society to truly feel the immense value brought by AI.
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。
