Tools That Changed How I Work — OpenAI Codex

Tools That Changed How I Work — OpenAI Codex

On April 16th, OpenAI shipped a capability expansion that turned Codex from a coding tool into a true agentic super app. Computer use, in-app browser, GPT Image 2.0, persistent memory, and a plugin ecosystem of ninety-plus integrations — layered on top of the desktop release from February. The practical effect was a step change. Codex stopped being a tool that helped me code and became the environment where most of my work now happens. In this episode, Chris Bradley walks through the seven capabilities that define Codex today, the prompt bake-off practice that drives his tool decisions, and where Codex pulls ahead — particularly on slide and visual work powered by GPT Image 2.0. He's also direct about the limits: Claude Code is still his primary IDE for heavy builds, NotebookLM still wins on source-grounded research, and both Codex and Claude have shown growing pains as the labs push harder at the frontier. The fourth episode in the Tools That Changed How I Work sub-series. The through-line across all four — NotebookLM, Antigravity, the Claude desktop app, and now Codex — is the same: match the tool to the work, and stay willing to move when the picture changes. Three things to try this week included.

Tools That Changed How I Work: Anthropic’s Claude Desktop (Code, Cowork & Chat)

The big three AI labs are all building superapps — but each is packaging them differently. Google blurs everything together inside Antigravity. OpenAI is consolidating into Codex. Anthropic took a third path: three distinct surfaces inside one Claude desktop app — Chat, Cowork, and Code. In this episode, Chris Bradley walks through how Anthropic's approach actually plays out in practice. When to use each surface. Why Claude Code desktop has become his go-to for vibe coding. How Cowork turns Claude from an AI chat into a real workspace that produces documents, spreadsheets, and decks. And why specialization with seamless switching solves the context-switching problem differently than pure consolidation does. Episode 6 in the Tools That Changed How I Work sub-series. Episode 7 goes deep on OpenAI Codex. Chris Bradley is CMO at Veritiv and head of Veritiv's AI Transformation Lab.

Tools That Changed How I Work: Google Antigravity

In Episode 1, I mentioned a competitive analysis I ran in thirty minutes that would have taken two weeks. A global network analysis — fifty-plus competitor locations, fifteen manufacturing hubs, proximity calculations, coverage gaps, and a full action plan for leadership. I never finished that story. The output wasn't a report. It was a functioning Node.js application with an interactive map my team still uses today. Google Antigravity didn't just run an analysis. It built a tool. In this episode, I break down exactly how that happened — and why Google Antigravity produces results that most agentic AI tools can't match. We cover the three capabilities that make it different: its multi-agent architecture, its deep Google ecosystem integration, and its best-in-class browser control — including the audit trail that lets you review every browser session after the fact. We also get into vibe coding — why Google Antigravity is my preferred environment for building knowledge worker applications, and what it means that non-technical professionals can now describe a dashboard or internal tool and have it built in a single session. Plus: the Handoff Test — a simple three-question framework for knowing when Google Antigravity is the right tool to reach for, and when something else is a better fit. What you'll learn: * Why Google Antigravity was built as a developer IDE — and why that matters for knowledge workers * How multi-agent parallel execution works and why it produces results in minutes that would otherwise take days * What browser control actually means, and why the audit trail separates Google Antigravity from every other tool doing it * How to vibe code a production-quality knowledge worker application without writing a single line of code * The Handoff Test: three questions that tell you when to reach for Google Antigravity AI Transformation Lab is hosted by Chris Bradley, CMO at Veritiv — a practitioner-focused series for leaders navigating the shift from generative to agentic AI.

Tools That Changed How I Work: Google NotebookLM

Most AI tools know the internet. NotebookLM knows your sources. That distinction sounds small. It isn't. When you're doing serious research — regulatory analysis, competitive intelligence, executive briefing prep — the difference between a tool that synthesizes the web and a tool that synthesizes your documents is the difference between fast and trustworthy. In this episode, Chris Bradley introduces Google NotebookLM and explains why source fidelity changes what's possible in knowledge work. Every claim cited. Every insight traceable. Every synthesis grounded in the materials you approved — nothing more, nothing less. Chris walks through five capabilities that make NotebookLM genuinely useful in practice: cross-document synthesis, Audio Overviews, Slide Decks, Infographics, and Data Tables. He also introduces the Source Test — a three-question framework for knowing when to reach for NotebookLM versus a general-purpose AI tool like Claude or ChatGPT. The episode opens with a real example: ten sources, two conflicting regulatory frameworks — the EU's PPWR and the Environmental Omnibus — and a structured cross-document analysis that would have taken two days done in minutes. This is the first episode in the Tools That Changed How I Work sub-series. The focus isn't product reviews. It's matching the right tool to the right work — and understanding exactly why that matching skill matters more than most people realize.

The AI Fluency Playbook: Three Habits That Keep You Ahead

Most people plateau with AI. Not because they lack skill — but because they mistake usage for fluency. They use the tool more, expect improvement to follow, and then wonder why the gap between them and the most effective people on their team isn't closing. This episode addresses that directly. The answer isn't a better training program. It's a better practice system. In Episode 3, Chris Bradley introduces a three-habit framework built for continuous AI fluency: the AI Pulse, a monthly calibration that keeps your mental model of AI's capabilities current; the Workflow Lab, a weekly thirty-minute session dedicated to examining and redesigning one workflow at a time; and the Workflow Pause, an in-the-moment reflex that catches automation opportunities before you execute on autopilot. The episode also shows how ODIF — the agentic direction framework from Episode 2 — maps directly onto workflow redesign, making every Lab session more rigorous and every output more immediately usable. The goal isn't to cover more ground. It's to build habits that compound — practices that run continuously.