Fast & Cool

OpenAI might be the hottest company in tech right now. The company is behind the popular large language-based model, ChatGPT, and also built the GPT-3, GPT-3.5, and GPT-4 models which several AI coding assistants use, such as GitHub Copilot.

Since it arrived last year, ChatGPT has been shipping new features at high speed

Gergely Orosz asked the question "What is the engineering culture that allows the ChatGPT team to iterate this fast?" to Evan Morikawa

You can find the interview at The Pragmatic Engineer

What resonates with me is that the decisions do not require any magic sauce and I will just mention two of them:

Operating like a small, independent startup: They want the atmosphere of an early-stage startup iterating towards product-market fit. ..So every team member was on-site, and rearranged seating to put people next to each other.

Created a tight integration with Research In most tech companies, there is a “classic” trio of teams referenced as “EPD”: Engineering, Product, and Design. These teams tend to collaborate heavily with one another, and cross-functional teams usually have members of engineering, product, and design within them. ChatGPT brought researchers into product teams.

Suddenly I realize that this is how we are in LATAM.

We are interconnected and very often we cross boundaries between different fields. Our unique genetic makeup may contribute to our success, but we may not fully realize its significance.

>