Original Title: Built to benefit everyone: our plan
Original Authors: Sam Altman, Jakub Pachocki, OpenAI

Every few generations, a new technology emerges that changes everything. Imagine the 1920s in America when electricity arrived in a remote town. Before the power lines were set up, daily life was constrained by physical limitations: carrying water, hand-washing clothes, storing food with ice, and once the sun set, most activities came to a halt. Electricity did not transform every household overnight, and the benefits were unevenly distributed.

But as electricity became more widespread, everyday life quietly transformed. The night lights extended the day; electric pumps, household appliances, and refrigerators eased many burdensome daily chores; radios brought news, music, and connection from hundreds of miles away into homes and public places. The initial purpose of electricity was to make people’s lives more convenient, but its deeper impact came from enabling more and more people to use it to create new possibilities. Over time, new possibilities emerged—machines and computers greatly accelerated progress in medicine, engineering, and many other fields.

By the end of the 20th century, human life expectancy had increased by over 20 years, and median income had approximately tripled in inflation-adjusted terms. Many of these achievements were largely due to improvements in healthcare, hygiene, and living standards, many of which were driven or accelerated by the widespread adoption of electricity and related technological advancements.

AI is now recreating this phenomenon. AI is poised to have extraordinary capabilities. But the focus is not on the technology itself, but on what people can do with it. It can help people understand a medical bill, learn a new skill, start a small business, care for elderly parents, comprehend a legal or financial decision, bring an idea to life, or even drive a scientific discovery.

The novelty of the night lights may soon fade, but what people decide to do with the light will not. And precisely because technology has always been a reliable pathway to prosperity, we believe AI should be open to everyone, allowing each person to use it as needed, anywhere, and in any way.

But such a future will not come about on its own. Disruptive technology can centralize power or decentralize it; it can make the lives of a few more convenient or expand opportunities for many. Our philosophy is based on one belief: AI should serve humanity—helping people achieve their goals, enhancing their capabilities, and spreading the benefits of this technology as widely as possible.

Our first commitment is to build AI to serve humanity. This means we aim to empower the masses rather than consolidate power in the hands of a few companies, governments, or individuals. We believe that a safer future is one where power is widely distributed, allowing more people around the world to participate in building a resilient ecosystem.

We are optimistic about AI because we believe it can enhance human abilities and well-being. But we also remain cautious about the risks. Powerful systems must ensure safety, align with human intent, and be subject to human oversight and control. OpenAI’s mission is to ensure that AGI benefits all of humanity. This means the systems we aim to build are meant to help people do more of what they choose, not to replace humans.

An entirely automated future is not what we want. It would be bleak and dangerous. AI should help people better themselves, not replace them. As AI systems become more capable, human roles will become more critical: setting direction, making trade-offs, exercising judgment, and bringing values, tastes, care, and responsibility into the work. In the future, a core mission for humanity will be to decide what is worth doing.

We believe that in the coming years, AI research conducted by AI will be a key determinant of the pace of progress. This is crucial because ‘alignment’ is itself a profound research challenge. To make rapid and deep progress, our researchers will need to rely on AI systems to test ideas, uncover errors, explore alternatives, and iterate side by side with us.

However, as technological progress accelerates, human judgment and public collaboration become even more critical, not less. The future should be shaped collectively by humans, institutions, and society, not just by companies building the most powerful systems. As cutting-edge AI development continues to progress, we expect that national and global coordination will become increasingly important.

We have long believed that ultimately an international organization should be established to coordinate the major AI powers to collectively mitigate catastrophic risks. Embedding cooperation and shared security as a standard is also a key part of the path forward—especially as challenges arising from both business competition and national rivalries are hard to avoid. One of the goals of such an organization should be to enable the world to take coordinated action, including slowing down development in cutting-edge areas when necessary to ensure that societal resilience, safety, and alignment keep pace with technological advancements.

Currently, OpenAI has three core goals: Build an Automated AI Researcher—an AI system that accelerates and increasingly automates the research process itself, while maintaining controllability, accountability, and close collaboration with humans. Our internal projection is that by March 2028, a significant portion of our research will be conducted collaboratively by AI systems and our own researchers. To make sufficient progress on aligned research, we believe AI needs to iterate with us. This will help us collectively navigate the transition to the “post-AGI world” and decide on the path to the future together.

Accelerate Economic Development—by accelerating scientific progress, boosting productivity, and driving economic growth, while ensuring that the benefits are widely shared. Everyone should have the opportunity to share meaningfully in the prosperity created by AI.

Provide Everyone on Earth with a Personal AGI—empowering them to benefit from one of humanity’s most transformative technologies in a way of their choosing. To achieve all this, we are entering OpenAI’s third act. OpenAI’s first act focused on AGI research. The second act began as our research started to get linked to the real world, and we became a product company: deploying our systems, learning from how people used them, and continuing to advance AGI progress in a safe and “mission-aligned” manner.

Today, we are entering the third phase. The economy is being reshaped around AI. The central question has become: How do we make advanced AI abundant, affordable, safe, practical, and easy to use, so that everyone and every organization can benefit from it? Frontier capabilities are just one part of the job; a bigger task is turning this capability into tools that people can really use to create and develop.

Most importantly, we believe that widespread empowerment will help lead to a better future. Human history shows that the concentration of power makes society fragile, while the broad sharing of power makes society more resilient, adaptable, and free. This is why democratization is important, as are security, privacy, affordability, an open ecosystem, and public oversight.

A better future with AI should not be a future where a handful of organizations control most of the capability and benefits. It should be a future that many people, numerous businesses, communities around the world, and countries can all build, benefit from, and hold power in. We believe that this transformation should belong to everyone. If done right, AI can be a cornerstone for increasing mass productivity, creativity, scientific progress, and economic opportunity, and we will also achieve our mission: to ensure that AGI benefits all of humanity.

[BlockBeats]