Inside OpenAI: Former Engineer Reveals Company Culture
Broken Communication Systems and 'No-Email Culture' Amid 3x Annual Growth

- •Segment founder Calvin French-Owen publicly shared OpenAI's company culture after one year with the organization.
- •OpenAI is experiencing communication system and organizational structure breakdowns as annual headcount surged from 1,000 to 3,000 employees.
- •The company features a Slack-centric culture without email and an influx of Meta talent, while maintaining a startup-like high-pressure environment.
MIT Graduate and Entrepreneur's Reality at OpenAI
Calvin French-Owen, co-founder of Segment who sold the company to Twilio for $3.2 billion (approximately 4.3 trillion won) in 2020, has publicly shared his candid experiences with OpenAI's company culture after leaving the organization.
The MIT graduate and successful entrepreneur joined OpenAI in May 2024, but admitted that despite his Silicon Valley credentials, he wasn't prepared for the shock that came with working at one of the world's leading AI research labs. After about a year with the company, his reason for leaving wasn't internal conflict but rather "a desire to return to startup life."
"There's plenty of hype about OpenAI, but very little ground truth about what it's actually like to work there," he wrote on his blog.
Chaos Born from 3x Annual Growth
During French-Owen's tenure, OpenAI's headcount exploded from 1,000 to 3,000 employees. This came as ChatGPT surpassed 500 million monthly active users, becoming the fastest-growing consumer product in history.
However, this explosive growth came with serious side effects. French-Owen explained: "When you grow this fast, everything breaks. How the company communicates, reporting structures, product launch processes, people management and organizational structure, hiring processes."
He particularly noted OpenAI's distinctive culture:
A Company Without Email: "Everything, and I mean everything, happens on Slack. There is no email. I probably received about 10 emails during my entire time there."
Meta Talent Pipeline: French-Owen mentioned that "there's a very active engineering talent pipeline moving from Meta to OpenAI." Recent reports indicate Meta is scouting OpenAI talent with signing bonuses worth $100 million (approximately 130 billion won), which OpenAI CEO Sam Altman has described as "insane."
Engineering Culture Resembling Early Meta
OpenAI's engineering culture shares many similarities with early Meta. There's a consumer-facing application, infrastructure development in progress, and extreme pressure for rapid releases.
Pros and Cons:
- Individuals can pursue their own ideas, but this leads to duplicate work
- French-Owen noted: "I saw about 6 different libraries for things like queue management or agent loops"
- Developer capabilities range from Google veterans to PhD graduates with minimal practical experience
- The main codebase, which heavily relies on Python, was described as "closer to a garbage pile"
Startup-Style High-Pressure Environment
OpenAI still maintains a startup-like culture. Amid the fast pace, high pressure, and constantly changing environment, employees focus on launching and improving products.
French-Owen's work on the Codex launch (a code-generation AI model) during his tenure exemplified this culture. He mentioned the intense pressure and rapid decision-making structure he experienced during product development.
[AI Analysis] The Light and Shadow of OpenAI's Growth
OpenAI's case demonstrates common challenges faced by rapidly growing AI companies.
Sustainability Issues: While a 200% annual growth rate is impressive in the short term, if organizational culture and systems can't keep pace, it's likely to lead to decreased productivity and talent attrition in the long run.
Intensifying Talent War: The talent battle between Meta and OpenAI is expected to become even fiercer. The $100 million bonus in particular signals an overheated AI talent market.
Cultural Identity Crisis: In transitioning from startup to large corporation, OpenAI faces the dilemma of maintaining its rapid execution culture while introducing systematic organizational management.
French-Owen's testimony reveals the reality behind OpenAI's glamorous external image, suggesting that the journey to becoming an AI industry leader is far from smooth.
댓글 (5)
댓글란이 과열되지 않았으면 합니다. 차분한 논의가 필요해요.
차분한 논의가 필요하다는 말에 공감합니다.
팩트에 기반한 냉정한 판단이 필요한 시점입니다.
이 사안은 신중하게 접근해야 한다고 봅니다.
균형 잡힌 시각이 필요하다는 데 동의합니다.
More in this series
More in AI & Tech

Reddit Considers Face ID to Block Bots While Maintaining Anonymity

China Reduces Hypersonic Missile Core Technology Simulation to 7 Days

BTS Gwanghwamun Concert: AI Network Prevents Communication Crisis for 40,000 Fans

Czech Drone Factory Fire Under Investigation for Terrorism

Trump Slams NATO Allies as 'Cowards' Over Strait of Hormuz Refusal

Google Unveils Gemini 3.1 Flash-Lite Optimized for High-Volume Processing
Latest News

이스라엘, 헤즈볼라 무기 통로 레바논 다리 공습
이스라엘군, 헤즈볼라 무기 통로 레바논 다리 공습

중동행 전세기 전쟁보험료 최고 7천500만원
중동행 전세기 전쟁보험료가 최고 5만달러(7천500만원)로 상승

이란 탄도미사일, 이스라엘 방어망 뚫고 160명 부상
이란 탄도미사일이 이스라엘 방공망을 통과해 160명 부상

Middle East Conflict Drives Manufacturing Outlook to 10-Month Low
The Korea Institute for Industrial Economics & Trade survey shows April manufacturing outlook PSI plummeted to 88, falling below baseline for the first time in 10 months.

Lee Jae-myung Administration Excludes Multi-Home Officials from Real Estate Policymaking
President Lee Jae-myung has ordered the exclusion of multi-home owning public officials from all real estate policy processes.

Southeast Asia Growth Forecasts Cut Amid Oil Price Surge, Threatening Korean Exports
Maybank Research has downgraded ASEAN-6's 2026 growth forecast from 4.8% to 4.5%.

Volkswagen CEO Says Germany Should Learn from China's Industrial Strategy
Volkswagen CEO stated that Germany should learn from China's systematic industrial planning approach.

BTS Tops March Artist Brand Reputation Rankings with First Full Group Comeback in 4 Years
BTS ranked first in the Korean Corporate Reputation Research Institute's March Artist Brand Reputation Rankings based on 99 million data points.