Skip to main content

What OpenAI Really Wants

OpenAI, co-founded by Sam Altman, is on a mission to build artificial general intelligence (AGI) that is safe for humanity. The company's journey has been marked by significant breakthroughs, challenges, and transformations.

In the early stages, OpenAI struggled to find a clear path towards its goal. However, the introduction of "transformer" architecture, which optimized the process of generating coherent text, marked a turning point. This led to the development of the generatively pretrained transformer (GPT-1), a model that outperformed everything that had come before in understanding language and generating answers.

OpenAI's subsequent iterations, GPT-2 and GPT-3, demonstrated even more impressive capabilities, raising concerns about potential misuse. This led to a temporary withholding of the full version of GPT-2, highlighting the company's commitment to safety.

However, the pursuit of AGI requires massive resources. OpenAI, initially a non-profit, had to transform into a "capped" profitable business to secure the necessary funding. The company's financial documents even stipulate a contingency for when AI disrupts our entire economic system.

Despite the progress, OpenAI's definition of AGI remains vague, with the determination left to the board. The company's leaders believe that if they reach their lofty profit cap, their products will likely have achieved AGI.

OpenAI's journey underscores the complexity and potential of AGI. As the company continues its quest, it remains committed to its mission of making AGI safe for humanity, even as it navigates the challenges of scaling and commercial demands.

Microsoft's CEO, Satya Nadella, and CTO, Kevin Scott, recognized the need for innovation in AI, leading to a significant investment in OpenAI. Despite having spent billions on AI research, Microsoft saw the value in OpenAI's focus on artificial general intelligence (AGI), a field that even the tech giant had not fully explored. The partnership has resulted in Microsoft investing $13 billion into OpenAI, demonstrating the high cost of being at the forefront of AI development.

Microsoft's investment has not only provided OpenAI with essential resources but also given Microsoft a non-controlling equity interest in OpenAI’s for-profit side, reportedly 49 percent. This deal has resulted in Microsoft having an exclusive license to commercialize OpenAI’s technology, and OpenAI committing to use Microsoft's cloud exclusively.

Despite some criticism, OpenAI's partnership with Microsoft has been transformative. Microsoft's product, Copilot, uses OpenAI's GPT to generate code on command, demonstrating the practical applications of AGI. However, this

The original article: https://www.wired.com/story/what-openai-really-wants/