on april 26, local time, openai ceo sam altman published a statement outlining the company’s core principles and long-term goals in the development of artificial intelligence. he noted that ai has the potential to bring about transformations far greater than those brought about by the steam engine or electricity, significantly enhancing individual capabilities and opportunities for self-fulfillment. however, this future is not inevitable—technological power could either become concentrated in the hands of a few superintelligent companies or be distributed more democratically among the public.
openai’s goal is to put true artificial general intelligence into the hands of as many people as possible, ensuring that agi benefits all of humanity, and to this end it has formulated five guiding principles. first is democratization: openai will resist the concentration of power in the hands of a few, ensuring that key decisions about ai are made through democratic processes and in accordance with principles of fairness. second is empowerment: openai believes that ai can help everyone achieve their goals, learn more, and lead happier lives, while building and deploying technologies in ways that minimize harm. third is inclusive prosperity: by making easy-to-use yet powerful ai systems available to everyone, living standards can be dramatically improved; this may require governments to explore new economic models and invest heavily in ai infrastructure. fourth is resilience: ai will introduce new risks related to biosafety, cybersecurity, and other areas, and openai will work with stakeholders to address these challenges, continuing an iterative deployment strategy that allows society to gradually adapt to each level of ai capability. fifth is adaptability: to prepare for an unpredictable future, we must continually adjust our strategies and remain transparent when our guiding principles evolve.
altman recalled that when they first released gpt-2, they were anxious about its societal impact; yet it was precisely that experience that led them to discover the crucial approach of iterative deployment. he stated that we are now entering a phase of rapid technological advancement with far-reaching consequences, where every decision deserves careful consideration. although we cannot get everything right, we will learn quickly and adjust our course, thereby driving technology to deliver more positive social outcomes.
