OpenAI, a pioneer in the realm of large models, boasts an impressive user base of hundreds of millions with its ChatGPT. Yet, the inner workings of this innovative company have long remained an enigma. Recently, an engineer who parted ways with OpenAI three weeks ago shed light on his year-long journey within the organization.

He portrayed a company culture marked by diversity, with teams embodying contrasting dynamics. Some teams operate akin to startups, characterized by swift and agile progress, while others adhere to more conventional methodologies. The Codex team, where he was part of, epitomized a frenetic pace, embodying the mantra of "action first, discussion later."

OpenAI fosters a bottom-up approach where engineers spearhead projects, with leadership advancements hinging on the ability to conceptualize and execute groundbreaking ideas. The company maintains a stringent confidentiality protocol while prioritizing AI safety as a paramount concern.

The engineer described OpenAI as an ambitious entity, pushing boundaries across multiple fronts and demonstrating remarkable flexibility in collaborative endeavors. However, he also acknowledged the reality of talent attrition, notably the departure of co-founder John Schulman and others, underscoring the challenges faced in sustaining such a dynamic environment.