EPISODE
Bill Shufelt, Sonya Huang & Pat Grady, Sean Frank, Alex Mashrabov
00:01:00 Bill Shufelt, a former hedge fund trader and ultramarathon runner, co-founded Athletic Brewing Company in 2017 to create high-quality non-alcoholic craft beers that cater to health-conscious consumers. In the conversation, he discusses his transition from finance to brewing, the development of a proprietary brewing process to produce flavorful non-alcoholic beers, and the company's rapid growth, including becoming the top-selling non-alcoholic beer in the U.S.
00:02:00 Sonya Huang, a partner at Sequoia Capital, discusses the current state of artificial general intelligence (AGI), emphasizing that AGI has arrived and urging founders to leverage existing technologies to address real-world problems. She highlights the evolution of AI capabilities, noting recent advancements in long-horizon agents that can persist and iterate towards outcomes without constant supervision. Huang also touches on the shift from product-led growth to agent-led growth, where AI agents autonomously select and utilize the best tools and services, potentially transforming traditional business models.
00:03:00 Sean Frank, CEO of Ridge Wallet, discusses the company's eagerness to advertise on OpenAI's ChatGPT platform, highlighting the high revenue per user from AI-driven traffic compared to traditional sources. He expresses optimism about the potential of ChatGPT ads to deliver high-intent customers and anticipates that the introduction of ads will further boost organic activity, leading to increased site visits and sales. Frank also touches on the evolving landscape of e-commerce, noting the importance of adapting to new advertising platforms and the potential impact on traditional website traffic.
00:04:00 Alex Mashrabov, CEO and co-founder of Higgsfield AI, discusses the emergence of AI-native social media agencies that create commercial ads end-to-end using AI, highlighting the efficiency and personalization this approach offers. He notes that Fortune 500 brands are increasingly hiring smaller agencies that utilize Higgsfield's platform to produce customized ads, enabling the creation of thousands of videos at a cost below $1 per video. Mashrabov also addresses the challenges posed by the growing size of video models, which are expected to exceed 100 billion parameters, and emphasizes the importance of integrating performance data to optimize content creation through reinforcement learning.