Summary of Key Points
This interview is a "truth or dare" session with Sam Altman, the founder of OpenAI. He discusses the current state of AI development (which has not yet reached the "iPhone moment" that would transform everything), the slower pace at which society is adapting to AI than expected, OpenAI's strategic decisions (such as discontinuing the Sora video generation tool and its own browser), the challenges of entrepreneurship and the learning process (from investor to founder), and his views on the future of AI (promoting scientific discoveries is more important than automation, and human connections will become more valuable). The overall message is that while AI is powerful, the pace at which it changes the world is not that fast. The key is to find the "soul" in human-computer interaction and to ensure that technology empowers ordinary people, rather than being controlled by a few.
1. AI Hasn't Reached the "iPhone Moment": What's Missing Is the "Magic" of Interaction
Altman argues that all the technical components for AI are in place, but what's lacking is the type of interaction that makes it impossible to return to old ways of doing things. For example, despite having an AI tool like Codex, he still copies and pastes emails manually—not because the technology is inadequate, but because the product is not smart enough to completely replace old habits. He uses the iPhone as an example: it wasn't the first phone with multi-touch functionality that made it successful, but rather the fact that it made everyone feel uncomfortable without it. What AI needs now is a user-friendly design that creates addiction and a product concept that redefines the relationship between humans and machines.
2. AI Is Advancing Faster Than Humans, but Society Adapts Like a "Slow Turtle"
The host mentions that the CEO of Shopify predicted a reshuffle of the business landscape by 2026, but Altman counters that the economy has significant inertia—people still prefer familiar brands and tools, and changes are slow to happen. He notes that when GPT-4 was released, he thought the software industry would be revolutionized, but nothing significant happened. He sees this inertia as a good thing, as it makes the transition smoother, but he also warns against overestimating the speed at which AI will change the world.
3. OpenAI's "Subtractions": Discontinuing Cool Products to Focus on General Intelligence
Altman reveals that OpenAI discontinued two promising products last year: the Sora video generation tool (which was too resource-intensive) and its own browser, Atlas (although excellent, it was deemed less important than developing general intelligence. His logic is that general intelligence is the "upstream" focus; once the models are smart enough and computing power becomes affordable, products and ecosystems will naturally emerge. OpenAI's goal is to create a platform where anyone can develop applications, allowing millions of new companies to use its APIs, rather than OpenAI dominating the entire economy.
4. The "Reverse Path" of Entrepreneurship: From Investor to Founder, the Power Law Is Key
Altman is one of the few in Silicon Valley who became an investor before starting a company. He says that investing taught him about the power law—the returns from a few successful projects can exceed the sum of returns from many smaller projects. Therefore, when starting a business, one should be willing to bet on "great ideas," even if it means sacrificing smaller, seemingly good ideas. He also emphasizes the importance of the YC (Young Entrepreneurs) ecosystem in enabling young and less experienced entrepreneurs to secure significant funding, which was crucial for OpenAI's success.
5. The Future of AI: Scientific Discovery Is More Important Than Automation, and Human Connections Will Be More Valuable
What excites Altman most about AI is its potential to drive scientific breakthroughs—discovering new physical laws, curing diseases, and solving complex mathematical problems. He believes these achievements are more meaningful than automating tasks, as they help humans understand the world better. He is confident that, no matter how advanced AI becomes, people will still value human interactions, such as face-to-face conversations and reading books. He also warns about two risks: AI becoming uncontrollable and the concentration of power in the hands of a few companies. OpenAI's mission is to empower people so that they can control their own future.
In Conclusion: Entrepreneurship Is About Finding Direction in Uncertainty
Altman recalls the early days of OpenAI, when a group of people gathered in an apartment with a whiteboard but didn't know what to do, so they just wrote papers and came up with ideas. He says everyone experiences moments of uncertainty, but the key is to gradually find direction through user feedback and alternative indicators (such as game rankings like Dota 2). Entrepreneurship doesn't begin with a perfect plan; it's about moving forward clumsily through chaos, just like OpenAI has grown from GPT-1 to where it is today, through iteration and persistence.
This interview shows that even AI pioneers have uncertainties, but their focus is always on how technology can serve humanity—not to replace humans, but to make them better assistants. It also reminds us that the speed at which technology changes the world ultimately depends on how quickly society adapts to it.