From ChatGPT to Instagram to Uber: The quiet architect behind the world’s most popular products | Peter Deng

I believe education will change significantly. I think about this often because I'm quite involved in my children's school, a role I took on after leaving OpenAI. What's fascinating is watching my son, who had the opportunity to test much of OpenAI's technology before its public release. I think I can safely say that's okay to share. When he was playing with ChatGPT and some of the latest models, at just nine years old, I could already see his brain rewiring. He started asking questions, and he had never heard the word 'prompt' before. However, his mind, exposed to this technology at an early age, unlocked certain abilities. I think you're able to think differently.

One hard lesson I learned at Uber is that sometimes your product actually doesn't matter. By "product," I mean the pixels you put on the screen or the features you build into your mobile app. At Uber, I learned this because, it pains me to say, the real product was the price and the estimated time of arrival (ETA). Often, people at tech companies view the product solely as its digital manifestation. However, from a holistic perspective, we humans consume the entirety of the product. This was one of the most impactful lessons I learned: sometimes the pixels don't matter as much as you think. While you should certainly fix bugs, doing so might not have as significant an impact as something more crucial to users, like price or ETA.

I remember taking a class in college that really stuck with me. It was called Language and Thought, taught by Herbert Clark. He had a thesis that blew my mind: language actually affects the way you think. Once I heard that, read his book, and listened to his lecture, I couldn't stop thinking about it because it rang so true. I grew up speaking Chinese, and I noticed that certain aspects of the Chinese language made me think differently when I learned English. There were also studies around this. I'm not entirely sure, but I believe in Russian, there are two different words for blue: one for a greenish-blue and another for a bright blue.

I think the first thing I would say is that going from 0 to 1 is different from going from 1 to 100. When you are in the 1 to 100 phase, which is where I spent a lot of my time, you might ask, "Is the 1 to 100 phase able to quadruple Instagram usage in two years?" That was very much a fun ride. There are many other examples from other companies. But when you go from 1 to 100, I think one of the things you really need to consider is that you have to plan your chess moves in advance. You have to truly think before you act and build systems that will allow you to grow sustainably faster. The 0 to 1 phase is about finding product-market fit. Then, when you get to 1 to 100, you're trying to ensure you can reach hyperscale as quickly as possible.

I have a saying, documented in what I call the PXTD API, which outlines how to work with me. A core principle within it, something I optimize for with everyone I support and hire, is this: if, in six months, I'm still telling you what to do, I've hired the wrong person. This philosophy has served me well on three levels. First, it reminds me, when hiring, to set a very high bar and not settle. If I do settle, it's likely that in six months, I won't be able to trust that person to work independently, and I'll still be directing them, which is not my goal or desire.

These are the five types I've found to be most enduring and, actually, the most distinct. I love how you put this, Len, when you talk about hiring the right people and how they're naturally motivated by different things. So, these are the five types we identified. Number one is the Consumer PM. This person is half designer, half product person, and truly obsessed with the details. They ask: Is it delightful? Is it crafted enough? They might exclaim: Oh, my goodness, this is three pixels off! I can't stand it. This is driving me nuts! Why is this so complex? These are the people you might think of as the quintessential PM, but that's just one type. Another type, on the other side, that we've discussed before, is the Growth PM. These individuals are half data scientist, half product person. They are wired to think numbers first and have a certain skepticism about them, especially the best ones. They'll say: I'm really skeptical. Show me the data. Let's run a test and prove it. I don't believe you. I start with these two in the framework because they are actually very different. One says: I have a vibe, I feel the vibe, this is better. The other says: No, I don't believe you. We should test this and prove it. That's a really healthy tension. I love having two people in a room debating that. It's great. We are going to get some good things done, and we're going to move the product forward.

One example, since we're talking about Instagram, is that we tried to build a camera-first app called Bolt. It didn't work. The premise was to reduce the pressure to share. You could open the camera, send things to people, get feedback, and go from there. The Instagram design team was top-notch; the app was well-designed and fast because the Instagram engineering team was excellent at making performant mobile apps. It had all the advantages we valued in Instagram, but when we launched it in New Zealand or Australia, it didn't work. We knew this because we looked at the retention graphs. Retention is the key indicator in any product you build, not the number of users or volume, but actual retention and cohort retention. If you plot the line and it asymptotes, you're in a good spot because it means people will continue to stay on the app over a period of time. That just didn't happen. The learning here was that even with the best team in the world and the best product taste, you can't predict what will succeed on the first try. Failure is okay; you just learn from it. Nobody wallowed over that. We actually had some technology built there that we were able to port over to the main app, which was very helpful. As the great American poet Sean Carter said, "It ain't a loss, it's a lesson." As a product person, it's important to see it not as a failure, but as a great learning experience. Now I'm that much smarter, and this is something I've just collected.