Alexander Embiricos: People love how thorough and diligent Codex is. It's not the fastest tool out there, but it is the most thorough and best at hard complex tasks.

Claire Vo: If you're a software engineer or somebody who's even just new to using some of these AI tools, where would you get started with Codex?

Alexander Embiricos: We're building it into a full software engineering teammate. One of the things that Codex is great at is simply answering questions. If you have a chat where Codex is producing these plans and you want to change something, it's actually really nice for the model if you just use the same chat to ask for changes to the plan and that way it has all this context in its head when it's ready to get going. This is a great starter flow that shows how flexible this platform is and how it can meet a bunch of people at a variety of levels of tasks.

Claire Vo: How is OpenAI using this for bigger features and bigger products?

Alexander Embiricos: We used Codex to build a Sora app for Android in 28 days and it immediately became the number one app in the app store.

Claire Vo: Welcome back to How I AI. I'm Claire Vo, product leader and AI obsessive here on a mission to help you build better with these new tools. Today we have Alexander Embiricos, product lead for Codex from OpenAI, and he's going to show us how you can get the most out of Codex. Whether you're a non-technical user trying to make changes to an existing codebase or want the power tips and tricks for getting the most out of it in the terminal. Let's get to it. This episode is brought to you by Brex. If you're listening to this show, you already know AI is changing how we work in real practical ways. Brex is bringing that same power to finance. Brex is the intelligent finance platform built for founders.

Claire Vo: Alex, thanks for joining How I AI. I'm excited about today's episode because we actually haven't seen a deep dive into Codex yet and we are going to get the expert take on how to get the most out of this tool. And I love that we're just going to dive in and do a zero to one hello world with Codex. So if you're a software engineer or somebody who's even just new to using some of these AI tools, where would you get started with Codex?

Alexander Embiricos: Codex is a coding agent. We're building it into a full software engineering teammate. But to get started, let's just talk about where most people use it, which is in their IDE. I happen to use VS Code. So I'll show you Codex in VS Code. You can also use it the Codex extension in any VS code fork like cursor, etc. So let's say that I just installed Codex from the VS Code extension marketplace.

Claire Vo: Yeah, let's do it. Let's truly zero.

Alexander Embiricos: All right. Okay. Truly zero to one. I mean, I'm not going to uninstall and log out, but we can pretend I did that. Right. So, what would happen then is I'm going to get this glyph here, which is the Codex extension. I have to click through some steps and log in. And so, in case you didn't know, Codex is included in your ChatGPT plan. So, you need a paid plan. So, if you have a plus plan, pro, business team or EDU, you can use Codex. And the limits are really generous.

Alexander Embiricos: Okay. So, let's say I have this thing up and truly zero to one. Let's pretend I just heard that this is a game, but I don't even know how to play this game. One of the things that Codex is great at is simply answering questions. I'm the product lead for Codex. So, I actually use Codex a lot for asking questions, probably more than most engineers do, cuz I don't want to bother engineers with silly questions. So, I might ask, how do I play this game? We just launched a new model today. So, I'm actually curious what model that used. 5.2. Cool. So, I'm just going to run npm rundev here as it's saying, boot up the server and let's take a look at the game.

Alexander Embiricos: What I have here is like a simple commander type game. I can move my character around. I can recruit troops. It looks like planting windmills is not implemented yet. And I've heard there's something wrong with the jump. Okay, that's way too high. So, let's get to work fixing some of these issues. So, what I can do here is I can just go and ask, let's just say that jump is way too big. Lower, please. And so for those of you who are new to coding agents, I mean this is pretty basic. I just wrote in plain natural language, plain English what the change that I wanted. And we can see Codex getting to work thinking up of a plan. Okay, I need to figure out how the jump works. I need to then reduce it and then I need to like make sure this whole thing works.

Claire Vo: I want to call out some stuff for folks listening. So, what you're basically showing is the process of starting with an existing codebase and you as a semi-technical user. What you're using Codex for is one, how do I even run this thing locally? And then two, you're setting up little parallel tasks. Do you feel like it's a better approach to set up parallel tasks and just have individual ones running or to do them in sort of a serial basis?

Alexander Embiricos: This totally depends. If I'm making changes, then I'm more likely to think about, okay, how likely is this change to conflict with another change? And typically, I'll either do one at a time or I'll use something called a work tree, which is, I guess, a bit more of an advanced concept. I'll use a work tree and I'll just send Codex off to do its work on a separate work tree.

Claire Vo: Let's take a minute to look at work trees because I think this is something that most folks that are new to these tools aren't really using particularly well.

Alexander Embiricos: Basically, if we're going to have Codex make changes, what I need then is I need two different copies of the codebase. But Git has this really nice affordance called a work tree which basically lets one Git instance track multiple copies of the codebase. So as a sort of a classic mammal, I am lazy and so I don't want to remember the commands for worktree. So typically the way that I would actually do this is I would just ask Codex to create work trees. I might say something like "Codex, create two new work trees off the main branch, one called French and one called German."

Claire Vo: I was just looking at this, as you open a new Codex instance in the CLI, you can type this dash dash and your first prompt right in one line. And this is like classic developer productivity. It's like I cannot be expected to press enter, wait, and then type my words.

Alexander Embiricos: Exactly. So now I have two folders. I might CD into French and I might run Codex: "translate the input field placeholder strings to French." So now I can open a new tab and I will now cd into the German tab instead and I'll run Codex and I'll say "translate the input placeholder string to German." And so now Codex can go work on both of these changes at the same time.

Claire Vo: But how is OpenAI using this for bigger features and bigger products?

Alexander Embiricos: Actually we just published a blog post about how we used Codex to build the Sora app for Android in 28 days and it immediately became the number one app in the app store. So you know four engineers 28 days number one app in the app store. A really cool sort of headline to take away here is that with coding agents it doesn't get easier but you just move way faster. They didn't go in and just say hey Codex build the Sora Android app and have it work. Actually they did try that and it didn't work. But instead what they did is they thought really hard about the architecture that they wanted the app to have. And they used a technique called planning.

Alexander Embiricos: Like for instance, we have a TypeScript SDK and maybe we want to write a Python SDK. I don't necessarily want a oneshot Codex on that. So I might say something like this: "Make a plan to build a Python SDK based off our TypeScript SDK." Some of our power users at OpenAI have gotten fairly opinionated about how they like their plans to work. Basically, what you do is you go to this blog post and you just copy this description. It's kind of like a meta plan. It's like, hey, when you plan, this is what a good plan looks like.

Alexander Embiricos: I have copied that into a markdown file plans.mmd. And so what I might actually do instead here is I might say "using plans.mmd make a plan." This is actually something that Codex is really good at. Like people love how thorough and like diligent Codex is. Here is the plan that it came up with. And so you can see it's about 120 lines. It's really important for me to read these and verify them. And so now what I can do is I could say "implement the plan." This would probably like a 30 minute to 1 hour task but I would be pretty confident in the results.

Claire Vo: First of all, I love that you said that it has a head. We see this a lot. This sort of like build a plan. You don't kind of get to bypass the architectural thinking of like how should this app be set up. But then once you have the kind of right-sized chunks of work, then you can use this planning strategy to then get what you're going to do all laid out and then have Codex execute it.

Alexander Embiricos: I kind of like the terms vibe coding and vibe engineering to be honest. When you're going to build a production app like Sora app that you know you have to scale, it's really important that you have that like architect think about the shape of the app. The bottlenecks are kind of like thinking about what code to write and then making sure that code is good. I think at the same time though Codex can be really powerful for those places where you just want to learn and you don't actually need like a scalable production ready app. So for instance we use Codex a lot for prototyping. The designers on Codex actually have a fully mostly vibecoded full prototype that they can just like design into with code.

Claire Vo: This episode is brought to you by Graphite, the AI powered code review platform. As developers adopt AI tools, code generation is accelerating. But code review hasn't caught up. Graphite fixes this with AI powered reviews and an automated merge queue.

Claire Vo: So I have to ask one question and then I do want to go to code review which is how do you decide between what needs a plan and what doesn't?

Alexander Embiricos: Obviously the harder the task the more likely you want to have a spec. But the lazy answer to your question is also it depends if I have time to wait for a plan or not. Codex has a feature called best of N where it'll just like do the same task four times and find out what works best.

Alexander Embiricos: You can ask it for code review. You could type slash review and basically ask it to start reviewing his code. And this is something that people really love. When you put the model in a certain mindset like hey you are a reviewer and you give it a different conversation context than the model that wrote the code you'll just get like even better critiques than you might get from a human engineer. Codex code review itself is enabled on pretty much every single repo at the company and reviews like pretty much all PRs.

Claire Vo: Let's get to our lightning round questions. You showed 5.2 model on SWE-bench Pro. Why does the interface to the model like Codex matter so much?

Alexander Embiricos: Two places. One is just the quality of model work and then the other is in the user experience. These models are changing all the time. I think we're shipping a new model like every two weeks recently. You need to know how to get the most out of the model. For example, we see these really interesting emerging behaviors where sometimes the model will decide to write a Python script to make many edits to code. By building both the harness and the model together we can be much more opinionated about what to do with the model.

Claire Vo: I spied with my little eye Atlas. Tell me your favorite Atlas use cases.

Alexander Embiricos: I have started just asking chat for everything instead. I tell it what decision I made because then it remembers. And so then the next answer I get is even better. My other sort of hot take reason to do this is I think it's important to be polite to AI. I just think it's important to be polite to everyone and I think that if you start not being polite to chat I think it can wear off on you and you just start not being polite to other people in your life.

Claire Vo: Could not agree more. It's to protect my own humanity. Lastly, what is your prompting technique to get it back on track?

Alexander Embiricos: Context is everything. I don't usually ask for things from the agent without giving context. I'll say like I want you to change this UI because we don't want people to be confused. I think PMs are the best prompters because we're used to not being the expert. Tip number one is give a lot of context and actually get really good at describing the level of ambiguity of your request. And then the second thing is like if that doesn't work, then I just start a new chat. Codex stores its conversations logs in a subfolder called sessions. You could just go say like, "Go read your previous session to understand what's going on."

Claire Vo: Alex, thank you so much. Where can we find you?

Alexander Embiricos: I am hiring PMs. So if you're interested, please apply on the job site and also hit me up on socials. You can find me on Twitter or the subreddit. Codex is awesome. So check it out!