OpenAI's Greg Brockman says it's not too late to build AI startups

For those contemplating a leap into the world of AI startups, the window of opportunity remains wide open. Greg Brockman, the co-founder and president of OpenAI, emphasized this point during a recent episode of the 'Latent Space' podcast. He noted that while it may seem like all viable ideas have been exhausted, the vastness of the market still holds immense potential. Brockman highlighted the importance of linking large language models with practical applications, especially in fields like healthcare. He encouraged aspiring entrepreneurs to consider all stakeholders involved and strategize on how to effectively integrate AI into existing frameworks. "There is so much fruit that is not yet picked, so go ahead and ride the GPT river," he asserted, motivating innovators to explore the uncharted territories of AI. He also cautioned against the temptation to create mere 'AI wrappers'—a term that disparages simplistic applications built on existing models. Instead, Brockman stressed the necessity of gaining a deep understanding of specific domains, cultivating expertise, and forging robust relationships within those industries. His remarks come amidst ongoing discussions in Silicon Valley about how aspiring AI founders can secure their future in an ever-evolving landscape. Last year, OpenAI's CEO, Sam Altman, warned that startups focusing solely on minor enhancements to existing models might find themselves overshadowed by larger companies. He referred to this trend as the "OpenAI killed my startup meme." In a related podcast, Mike Krieger, co-founder of Instagram and now chief product officer at Anthropic, shared insights for startups aiming to thrive alongside AI titans. He advised that those with specialized knowledge in niches such as law or biotechnology, along with strong customer engagement, are better positioned to withstand competition from larger AI firms. Krieger also encouraged experimentation with new AI interfaces, even if they initially seem unconventional. Reflecting on the challenges, he admitted that he does not envy those looking to establish themselves in the AI sector. "Maybe that's part of the reason why I wanted to join a company rather than start one," he concluded, underscoring the complexities of the current entrepreneurial landscape in AI.