It’s code red at OpenAI. Just over three years after the company upended the tech world with the launch of ChatGPT—initially presented as a “low-key research preview”—the AI pioneer is now scrambling to keep pace with a rapidly intensifying competitive landscape. The once-dominant player in generative AI is facing a new reality, one where rivals like Google with Gemini, Anthropic, and others are pushing forward with faster releases, broader capabilities, and stronger integration across platforms. CEO Sam Altman recently sent an internal memo to OpenAI staff declaring a “code red,” urging the team to refocus on the company’s most critical products. The message was clear: OpenAI must accelerate innovation and sharpen its priorities to avoid falling behind. The challenge isn’t just about keeping up—it’s about proving that ChatGPT can evolve beyond a powerful language tool into something more transformative. But what does “making ChatGPT better” actually mean in 2024? The early promise of AI as a universal assistant capable of reasoning, planning, and understanding context has not fully materialized. While models have grown more fluent and capable, they still struggle with consistency, factual accuracy, and deep comprehension. The core question now is whether large language models (LLMs) are the right foundation for the next leap in artificial intelligence—or if they’re hitting a ceiling. On this episode of The Vergecast, David and Nilay dive into the current state of ChatGPT and the broader AI industry. They explore whether LLMs are fundamentally limited by their reliance on language patterns rather than true understanding. Even if we accept that this is the peak of the technology for now, what new products could still be built? From AI agents that manage schedules and workflows to integrated systems that operate across apps and devices, the potential remains vast—but the path forward is unclear. Before turning to AI, the hosts cover a few other major tech developments. Samsung unveiled its latest foldable innovation: a trifold smartphone. The device is either a bold leap in mobile design or a gimmick with limited practicality—opinions are divided. Meanwhile, Apple continues to see leadership changes, with design chief Alan Dye stepping down to launch a design studio at Meta. Some see this as a loss for Apple, questioning whether the company is losing its creative edge. Others argue that such turnover is natural in a large, complex organization. (A follow-up announcement of additional executive changes came shortly after the recording, fueling further debate.) As the AI race heats up and the boundaries of what’s possible remain uncertain, one thing is clear: the era of AI dominance isn’t just about who builds the best model. It’s about who can build the best experiences, integrate AI meaningfully into daily life, and maintain trust and reliability. OpenAI’s code red may be a wake-up call—but it’s also a reminder that the future of AI is still being written.