OpenAI Product Manager Ty Geri on the Future of Prompting

Geri’s approach mirrors that of industry veterans like Andrej Karpathy, who advocates for messy, vocalized sessions to help clear mental fog. By dictating his thoughts into ChatGPT during long, rambling sessions, Geri parses through complex problems to identify the core of his logic. Despite these specific habits, Geri argues that the ultimate goal of AI development is to remove the need for such elaborate prompting altogether. He acknowledges that OpenAI must still improve its writing quality, aiming for a future where the model functions intuitively without requiring specialized user tricks.

This shift in interaction fundamentally changes the role of a product manager. In the past, PMs relied heavily on data scientists to bridge the gap between concept and execution. Today, tools like vibe-coding allow managers to ship code directly, eliminating the technical excuses that once defined the profession. Geri emphasizes that while the job description remains centered on user advocacy, the barrier to entry for technical tasks has evaporated. His advice to peers is clear: the ability to understand and execute is no longer restricted by a lack of formal engineering experience. For those at OpenAI, the mission is to leverage these fluid, dynamic capabilities to deliver the best possible user experience, regardless of traditional departmental boundaries.