OpenAI CEO Sam Altman published an article entitled "The Gentle Singularity" on his personal blog yesterday (11), detailing his vision for the future development of artificial intelligence (AI). Altman believes that humanity has crossed the technological "event horizon" and is moving towards the era of digital superintelligence, and that this change will unfold in a smooth rather than abrupt way. (Synopsis: Young people use ChatGPT as a life guide!) OpenAI CEO Sam Altman: Over 35 years old can't keep up at all) (Background added: Sam Altman looks at the future of AI from a novice father: humanoid robots are coming, are you prepared?) OpenAI CEO Sam Altman published an article titled "The Gentle Singularity" yesterday (11), detailing his vision for the future development of artificial intelligence (AI). Altman believes that humanity has crossed the technological "event horizon" and is moving towards the era of digital superintelligence, and that this change will unfold in a smooth rather than abrupt way. In this blog post, Altman makes a few core points, including: Current AI systems have surpassed human capabilities in many ways, and hundreds of millions of people rely on them every day to complete important tasks, showing that AI is deeply integrated into daily life. He predicts that AI agents capable of performing cognitive work will emerge in 2025, AI systems may generate new insights in 2026, and physical robots performing real-world tasks in 2027. Altman emphasized that the 2030s will usher in a great abundance of intelligence and energy, promote scientific progress and productivity leaps, and greatly improve the quality of life. He pointed out that the recursive self-improvement of AI's acceleration of its own research, as well as the self-reinforcing cycle of automating the supply chain of data centers and robots, will bring smart costs closer to the cost of electricity, enabling a future where "smart is too cheap to meter." However, he also acknowledges challenges, including the fair distribution of AI alignment issues (ensuring AI is in humanity's long-term interest) and superintelligence. He suggested that after solving the alignment problem, we should focus on making superintelligence cheap and widely available, avoiding excessive concentration in the hands of a few individuals or companies, in order to promote social adaptation and equity. The following is the full text of Sam Altman's blog post "The Gentle Singularity." A gentle singularity While some jobs may disappear, new opportunities and wealth will emerge We have crossed the event horizon and take-off has begun. Humanity is on the verge of creating digital superintelligence, which is far less bizarre than imagined, at least so far. Robots haven't been seen walking around on the streets, and most people haven't talked to artificial intelligence (AI) all day. People still die of illness, we still can't easily travel to space, and there are many things in the universe that we don't understand. However, we have recently built systems that are smarter than humans in many ways, and that can dramatically increase user output. The most unlikely part is a thing of the past; Let's build scientific insights into systems like GPT-4 and o3, hard-won, but will take us far. AI will contribute to the world in many ways, but the quality gains that AI will bring to faster scientific progress and increased productivity will be enormous; The future may be far brighter than the present. Scientific progress is the biggest driver of overall progress; It's exciting to imagine how much more we can have. In a significant sense, ChatGPT is already more powerful than any human who has ever existed. Hundreds of millions of people rely on it every day for increasingly important tasks; A small new feature can have a huge positive impact; A small deviation multiplied by hundreds of millions of people can have a huge negative impact. 2025 will see the arrival of agents who can perform real cognitive work; Writing computer code will change forever. In 2026, there may be systems that uncover new insights. In 2027, there may be robots that can perform tasks in the real world. More people will be able to create software and art. But the world is also in greater demand for both, and experts may still be far superior to novices, as long as they embrace new tools. Overall, by 2030, the amount of work one person can accomplish far more than in 2020 will be an amazing change, and many people will find ways to benefit from it. In the most important aspects, the 2030s may not be much different from the present. People still love their families, express creativity, play games, and swim in the lake. But in some ways that are still very important, the 2030s could be very different from any previous era. We don't know how far intelligence can surpass humans, but we're about to find out. In the 2030s, intelligence and energy, creativity and the ability to realize it, will become extremely rich. These two have long been fundamental constraints to human progress; With plenty of intelligence and energy (and good governance), we can theoretically have anything. We've coexisted with amazing digital intelligence, and after the initial shock, most of us have become quite adaptable. Soon, we went from marveling that AI can generate a beautiful text to wondering when it can write a beautiful novel; From marveling at its life-saving medical diagnosis to wondering when it will develop a cure; From marveling that it can create small computer programs to wondering when it will be able to create a completely new company. This is the course of the singularity: miracles become routine, and then basic requirements. We've heard scientists say they're two to three times more productive than they were before AI. Advanced AI is compelling for many reasons, but perhaps the most important is that we can use it to accelerate AI research. We may discover new computing substrates, better algorithms, and even more unknowns. If we could complete ten years of research in a year, or even a month, the rate of progress would obviously be very different. From now on, the tools we've built will help us gain more scientific insights and help us build better AI systems. Of course, this is not the same as AI systems updating their code completely autonomously, but this is still the prototype of recursive self-improvement. There are other self-reinforcing cycles at play. The creation of economic value has set in motion an ever-expanding infrastructure flywheel to run these increasingly powerful AI systems. Robots that can build other robots (and to some extent, data centers that can build other data centers) are not far off. If we had to make the first million humanoid robots in the traditional way, but then they could run the entire supply chain, mining and refining minerals, driving trucks, operating factories, etc., to build more robots, and then build more chip manufacturing plants, data centers, etc., the rate of progress would obviously be very different. As data center production is automated, the cost of intelligence should eventually approach the cost of electricity. (People often wonder how much energy ChatGPT consumes in a single query; The average query uses about 0.34 watt-hours, which equates to a little more than a second of oven operation, or a few minutes of high-efficiency light bulb operation. A query also uses about 0.000085 gallons of water, which is about one-fifteenth of a teaspoon. The pace of technological progress will continue to accelerate, and so will the ability of humans to adapt to almost anything. Some challenges can be tough, such as...
The content is for reference only, not a solicitation or offer. No investment, tax, or legal advice provided. See Disclaimer for more risks disclosure.
Sam Altman's Real Observations on AI for the Future: The Singularity Will Not Explode, But Will Quietly Devour Everything
OpenAI CEO Sam Altman published an article entitled "The Gentle Singularity" on his personal blog yesterday (11), detailing his vision for the future development of artificial intelligence (AI). Altman believes that humanity has crossed the technological "event horizon" and is moving towards the era of digital superintelligence, and that this change will unfold in a smooth rather than abrupt way. (Synopsis: Young people use ChatGPT as a life guide!) OpenAI CEO Sam Altman: Over 35 years old can't keep up at all) (Background added: Sam Altman looks at the future of AI from a novice father: humanoid robots are coming, are you prepared?) OpenAI CEO Sam Altman published an article titled "The Gentle Singularity" yesterday (11), detailing his vision for the future development of artificial intelligence (AI). Altman believes that humanity has crossed the technological "event horizon" and is moving towards the era of digital superintelligence, and that this change will unfold in a smooth rather than abrupt way. In this blog post, Altman makes a few core points, including: Current AI systems have surpassed human capabilities in many ways, and hundreds of millions of people rely on them every day to complete important tasks, showing that AI is deeply integrated into daily life. He predicts that AI agents capable of performing cognitive work will emerge in 2025, AI systems may generate new insights in 2026, and physical robots performing real-world tasks in 2027. Altman emphasized that the 2030s will usher in a great abundance of intelligence and energy, promote scientific progress and productivity leaps, and greatly improve the quality of life. He pointed out that the recursive self-improvement of AI's acceleration of its own research, as well as the self-reinforcing cycle of automating the supply chain of data centers and robots, will bring smart costs closer to the cost of electricity, enabling a future where "smart is too cheap to meter." However, he also acknowledges challenges, including the fair distribution of AI alignment issues (ensuring AI is in humanity's long-term interest) and superintelligence. He suggested that after solving the alignment problem, we should focus on making superintelligence cheap and widely available, avoiding excessive concentration in the hands of a few individuals or companies, in order to promote social adaptation and equity. The following is the full text of Sam Altman's blog post "The Gentle Singularity." A gentle singularity While some jobs may disappear, new opportunities and wealth will emerge We have crossed the event horizon and take-off has begun. Humanity is on the verge of creating digital superintelligence, which is far less bizarre than imagined, at least so far. Robots haven't been seen walking around on the streets, and most people haven't talked to artificial intelligence (AI) all day. People still die of illness, we still can't easily travel to space, and there are many things in the universe that we don't understand. However, we have recently built systems that are smarter than humans in many ways, and that can dramatically increase user output. The most unlikely part is a thing of the past; Let's build scientific insights into systems like GPT-4 and o3, hard-won, but will take us far. AI will contribute to the world in many ways, but the quality gains that AI will bring to faster scientific progress and increased productivity will be enormous; The future may be far brighter than the present. Scientific progress is the biggest driver of overall progress; It's exciting to imagine how much more we can have. In a significant sense, ChatGPT is already more powerful than any human who has ever existed. Hundreds of millions of people rely on it every day for increasingly important tasks; A small new feature can have a huge positive impact; A small deviation multiplied by hundreds of millions of people can have a huge negative impact. 2025 will see the arrival of agents who can perform real cognitive work; Writing computer code will change forever. In 2026, there may be systems that uncover new insights. In 2027, there may be robots that can perform tasks in the real world. More people will be able to create software and art. But the world is also in greater demand for both, and experts may still be far superior to novices, as long as they embrace new tools. Overall, by 2030, the amount of work one person can accomplish far more than in 2020 will be an amazing change, and many people will find ways to benefit from it. In the most important aspects, the 2030s may not be much different from the present. People still love their families, express creativity, play games, and swim in the lake. But in some ways that are still very important, the 2030s could be very different from any previous era. We don't know how far intelligence can surpass humans, but we're about to find out. In the 2030s, intelligence and energy, creativity and the ability to realize it, will become extremely rich. These two have long been fundamental constraints to human progress; With plenty of intelligence and energy (and good governance), we can theoretically have anything. We've coexisted with amazing digital intelligence, and after the initial shock, most of us have become quite adaptable. Soon, we went from marveling that AI can generate a beautiful text to wondering when it can write a beautiful novel; From marveling at its life-saving medical diagnosis to wondering when it will develop a cure; From marveling that it can create small computer programs to wondering when it will be able to create a completely new company. This is the course of the singularity: miracles become routine, and then basic requirements. We've heard scientists say they're two to three times more productive than they were before AI. Advanced AI is compelling for many reasons, but perhaps the most important is that we can use it to accelerate AI research. We may discover new computing substrates, better algorithms, and even more unknowns. If we could complete ten years of research in a year, or even a month, the rate of progress would obviously be very different. From now on, the tools we've built will help us gain more scientific insights and help us build better AI systems. Of course, this is not the same as AI systems updating their code completely autonomously, but this is still the prototype of recursive self-improvement. There are other self-reinforcing cycles at play. The creation of economic value has set in motion an ever-expanding infrastructure flywheel to run these increasingly powerful AI systems. Robots that can build other robots (and to some extent, data centers that can build other data centers) are not far off. If we had to make the first million humanoid robots in the traditional way, but then they could run the entire supply chain, mining and refining minerals, driving trucks, operating factories, etc., to build more robots, and then build more chip manufacturing plants, data centers, etc., the rate of progress would obviously be very different. As data center production is automated, the cost of intelligence should eventually approach the cost of electricity. (People often wonder how much energy ChatGPT consumes in a single query; The average query uses about 0.34 watt-hours, which equates to a little more than a second of oven operation, or a few minutes of high-efficiency light bulb operation. A query also uses about 0.000085 gallons of water, which is about one-fifteenth of a teaspoon. The pace of technological progress will continue to accelerate, and so will the ability of humans to adapt to almost anything. Some challenges can be tough, such as...