ChatGPT Codex Voice + browser + Sites: an expert’s AI workflow | Nick Baumann (OpenAI)

August 3, 2026

AI Summary

5 min read

Nick Baumann, a developer experience lead at OpenAI, walks Claire Vowe through the most potent, underused capabilities of ChatGPT Codex and ChatGPT Work. The conversation centers on three workflows: using voice as a high-bandwidth delegation layer, building and deploying live websites from a mobile phone, and offloading the tedium of video editing to an AI agent. Across each, a clear psychology emerges—the best AI workflows are not about crafting perfect prompts but about lowering the friction to start, letting the model manage its own context, and treating the interaction like managing a capable but forgetful assistant.

Voice as a delegation layer, not a dictation tool

Baumann’s core insight is that the new voice mode in Codex is not just a faster way to type. It is a fundamentally different interaction paradigm. “If anybody has been lucky enough to have an assistant, this is how it gets done,” he says. The key mechanism is that voice can now spin up and manage multiple independent threads in the background. Baumann demonstrates this live: he triggers voice with a hotkey, asks it to book flights and a hotel for a Paris off-site, and then, mid-conversation, delegates a separate expense report task to a new thread. The model forks the work, uses existing browser sessions and plugin connections (Gmail, Navan, calendar), and reports back when ready. The user never has to swi

Continue reading the full summary in the app — free to try.

Read Full Summary →

Free • No credit card required

Listen to Audio Summary Open in App

Never miss an episode of How I AI

Get every new episode summarized in your inbox — free, ~5 minutes to read.

No spam. Unsubscribe anytime.

What you'll learn

1 (00:00) **Voice + Codex: The New Delegation Layer** - Nick demonstrates how voice on the Codex app lets you delegate complex multi-step tasks by talking naturally, just like you would to a human assistant.
2 (03:47) **The Meta Capabilities Most Users Miss** - Codex can create new threads, fork existing ones, and let the AI manage its own context—features that are powerful but not obvious.
3 (05:52) **Chat GPT Work on Mobile: Always-On Agent** - The Chat GPT work experience on mobile is essentially the Codex harness with all plugins, accessible from your phone.
4 (06:56) **Live Demo: Voice-Orchestrated Travel Booking** - Nick walks through a real-time voice session where Codex books flights, checks calendar, and expenses a purchase—all without typing.
5 (10:52) **Why Voice Lowers the Activation Barrier** - Voice-to-voice interaction solves “blank chat window syndrome” and makes delegation feel natural.
6 (13:28) **Hot Take: Latency vs. Intelligence in Voice** - Nick and Claire debate which matters more for voice experience, landing on latency as the critical differentiator.
7 (14:37) **Building a Live Site with Chat GPT Sites** - Nick builds a private website for the How I AI podcast live, using natural language and deploying it in minutes.

+ Full timestamped outline available in the app

Show Notes

Nick Baumann is on the Developer Experience team at OpenAI, where he spends his days building with, testing, and communicating the capabilities of ChatGPT Codex and ChatGPT Work. In this episode, Nick walks me through several features that have launched or evolved recently: the new voice interface with its screen-reading orb, the Heartbeats automation system in ChatGPT Work on mobile, the live ChatGPT Sites deployment feature, and his personal use case for AI-assisted UGC video editing.

What you’ll learn:

How two-person voice chat works
How Heartbeats work
How to build and deploy a live website with ChatGPT Sites
How to delegate a flight search, hotel booking, and expense report to Codex in a single voice conversation without opening a single app manually
Why ChatGPT Work on mobile is the most underutilized AI workflow for people already using the ChatGPT app
How to use a custom UGC Video plugin to feed 50 raw clips into ChatGPT, let it pull transcripts, pick the best takes, and assemble a finished vertical video overnight

—

Brought to you by:

Bolt.new—Turn your idea into a real product

Hyperagent—Deploy fleets of agents that handle real work

—

In this episode, we cover:

(00:00) Introduction to Nick Baumann

(02:56) What’s new in Codex

(05:40) ChatGPT Work and Heartbeats

(06:40) Live Codex voice demo

(13:25) Latency vs. intelligence

(14:36) Quick recap