Three weeks in the past, an engineer named Calvin French-Owen, who labored on one in all OpenAI’s most promising new merchandise, resigned from the corporate.
He simply revealed a fascinating blog post on what it was prefer to work there for a 12 months, together with the sleepless dash to construct Codex. That’s OpenAI’s new coding agent that competes with tools like Cursor and Anthropic’s Claude Code.
French-Owen stated he didn’t depart due to any “drama,” however as a result of he needs to get again to being a startup founder. He was a co-founder of buyer knowledge startup Section, which was bought by Twilio in 2020 for $3.2 billion.
A few of what he revealed concerning the OpenAI tradition would shock nobody, however different observations fight some misconceptions concerning the firm. (He couldn’t be instantly reached for remark.)
Quick development: OpenAI grew from 1,000 to three,000 folks within the 12 months he was there, he wrote.
The LLM mannequin maker definitely has causes for such hiring. It’s the fastest-growing shopper product ever, and its opponents are additionally rising quick. In March, it stated that ChatGPT had over 500 million active users and climbing shortly.
Chaos: “Every little thing breaks whenever you scale that shortly: easy methods to talk as an organization, the reporting constructions, easy methods to ship product, easy methods to handle and manage folks, the hiring processes, and many others.,” French-Owen wrote.
Like a small startup, folks there are nonetheless empowered to behave on their concepts with little to no pink tape. However that additionally signifies that a number of groups are duplicating efforts. “I need to’ve seen half a dozen libraries for issues like queue administration or agent loops,” he provided as examples.
Coding talent varies, too, from seasoned Google engineers who write code that may deal with a billion customers, to newly minted PhDs who don’t. This, coupled with the versatile Python language, signifies that the central code repository, aka “the back-end monolith,” is “a little bit of a dumping floor,” he described.
Stuff continuously breaks or can take extreme time to run. However prime engineering managers are conscious of this and are engaged on enhancements, he wrote.
“Launching spirit”: OpenAI doesn’t appear to know but that it’s a large firm, proper right down to operating solely on Slack. It feels very very similar to move-fast-and-break-things Meta in its early Fb years, he noticed. The corporate can also be filled with hires from Meta.
French-Owen described how his senior crew of round eight engineers, 4 researchers, two designers, two go-to-market employees, and a product supervisor constructed and launched Codex in solely seven weeks, begin to end, with nearly no sleep.
However launching it was magic. Simply by turning it on, they obtained customers. “I’ve by no means seen a product get a lot speedy uptick simply from showing in a left-hand sidebar, however that’s the ability of ChatGPT.”
Secretive fishbowl: ChatGPT is a extremely scrutinized firm. This led to a tradition of secrecy in an try and clamp down on leaks to the general public. On the similar time, the corporate watches X. If a put up goes viral there, OpenAI will see it and, probably, reply to it. “A buddy of mine joked, ‘this firm runs on twitter vibes,’” he wrote.
Largest false impression: French-Owen implied that the most important false impression about OpenAI is that it isn’t as involved about security appropriately. Actually lots of AI security people, together with former OpenAI employees, have criticized its processes.
Whereas there are doomsayers worrying about theoretic dangers to humanity, internally there’s extra deal with sensible security like “hate speech, abuse, manipulating political biases, crafting bio-weapons, self-harm, immediate injection,” he wrote.
OpenAI isn’t ignoring the long-term potential impacts, he wrote. There are researchers taking a look at them, and it’s conscious that tons of of tens of millions of persons are utilizing its LLMs as we speak for every little thing from medical recommendation to remedy.
Governments are watching. Rivals are watching (and OpenAI is watching opponents in return). “The stakes really feel actually excessive.”