5 interesting quotes from Sam Altman's livestream

In a recent livestream, OpenAI CEO Sam Altman engaged with an audience, sharing his ambitious goals for the future of artificial intelligence. He expressed a clear intention: by March 2028, the company aims to develop a true automated AI researcher. During the session, Altman addressed various topics, including OpenAI's evolving definition of artificial general intelligence (AGI) and the company's commitment to providing flexibility for adult users of its technologies. While discussing past missteps, he acknowledged a faux pas regarding a controversial example he had used when discussing content guidelines. "I wish I had used an example other than erotica," he remarked, highlighting the need for a nuanced understanding of content differentiation. Altman also emphasized the importance of infrastructure in AI development, revealing OpenAI's aspirations to construct a facility capable of generating one gigawatt of computing power per week. This ambitious plan could significantly enhance the company’s capabilities, provided that research continues to progress alongside consumer demand. He noted, "We're not committed to this yet, but we are having conversations about it. Our aspiration is to build an infrastructure factory to meet these demands efficiently." The Q&A session also touched on the status of older AI models, with Altman reassuring users that OpenAI has no immediate plans to phase out the widely used GPT-4. He acknowledged the strong user attachment to this model and stated, "We understand that it's a product that some of our users really love." Questions surrounding age verification for ChatGPT users were also prominent. Altman indicated that verified adult users would enjoy greater flexibility, although he firmly stated, "We're not going to do the equivalent of selling heroin or whatever, even if you sign a liability." This livestream marked a significant moment for OpenAI, coinciding with the completion of a restructuring and a definitive agreement with Microsoft. As the discussion unfolded, it became clear that the pursuit of AGI remains a central focus for OpenAI, with the company seeking to clarify its goals rather than get bogged down in the debate over definitions. Altman concluded with a compelling vision: the ultimate aim is to develop an automated AI research intern by September 2026, paving the way for future advancements.