Welcome, humans.

So let’s talk about that ClickUp layoff letter. A bit of time has passed for this one, but it’s worth reviewing now that the dust has settled.

ICYMI, ClickUp CEO Zeb Evans said the company cut headcount by 22% while the business is “the strongest it’s ever been,” then laid out a pretty stark theory of the AI workplace: the winners will reorganize around builders, agent managers, and front-liners. We prefer the slop cannons and adults meme, but okay let’s hear him out. 

His version looks like this: the best engineers direct agents and review their work. Product managers and designers merge into product builders who use agents for research, prototyping, and iteration. The people who automate their own jobs become system owners. And customer-facing humans spend more time with customers because everything around the meeting gets automated.

The sharpest point of his vision is that more AI output can create more drag (Dan Shipper’s fantastic essay said the same). A company can celebrate 500% more pull requests, then quietly bury its best engineers under review work. This is why so-called aimless tokenmaxxing is basically just burning a bunch of money. 

So the “100x org” idea is less about everyone generating more stuff (more better!), and more about putting the right people in charge of deciding what should exist… then letting them cook. 

For his part, Dan simplifies this organizational change to pirates and architects. But of course, it’s a little more sophisticated than that.

Anyway, congrats to everyone who learned prompt engineering (2023-2026). The next skill tree to master is agentic reorganization… best of luck to us all.

Here’s what happened in AI today:

• 🐱 OpenAI is turning ChatGPT into one app that does everything

• 📰 Banks are cutting junior analyst classes by up to two-thirds

• 📰 Microsoft launched seven in-house AI models, ditching its OpenAI exclusivity

• 📰 xAI secretly trained on Claude outputs, kept going after getting caught

Hey: Want to reach 700,000+ AI-hungry readers? Advertise with us! 

P.S: Love robots? We’re starting a new robotics newsletter! Sign up early here.

😸 OpenAI Admitted Its Product Strategy Was Broken. Now It's Building One App to Fix It.

You probably use ChatGPT for some things, Google for others, and a separate coding tool for others still. OpenAI built products that way too: ChatGPT over here, Codex (their AI coding platform) over there, and an AI browser called Atlas somewhere else. By late 2025, Sam Altman had seen enough. He declared an internal "code red" and told the company to stop the sprawl.

The result is coming to your desktop soon.

Here's what happened:

• OpenAI confirmed it is merging ChatGPT, Codex, and its Atlas browser into a single desktop "superapp," first reported by the Wall Street Journal in March and now entering its rollout phase

• The Financial Times report added the commercial layer: the redesign will steer users toward partner services like Canva and Booking.com, with coding tools, AI agents, and third-party integrations baked in

• Greg Brockman, OpenAI's co-founder and president, is permanently leading product strategy for the unified platform; Fidji Simo (OpenAI's CEO of Applications) is leading the commercial push

• Simo's internal memo was blunt: "Fragmentation has been slowing us down and making it harder to hit the quality bar we want"

• The mobile app is not changing; this is a desktop-first overhaul targeting professional and enterprise users

• OpenAI filed a confidential S-1 (the document companies submit before going public) with the SEC days after the restructuring was announced, at a reported valuation above $300 billion

Why this matters:

The "superapp" model, popularized by WeChat in China and Grab in Southeast Asia, means one platform handles everything: chat, search, productivity, payments, commerce. OpenAI is betting that the AI equivalent is possible, and that ChatGPT is the right container for it.

This is also a direct shot at Anthropic. Claude Code hit $2.5 billion in annualized revenue by February 2026 and has been eating into OpenAI's lead with developers. Merging ChatGPT and Codex into one platform, with enterprise sales led by Simo, is OpenAI's answer: stop competing product by product and just own the whole workflow.

The IPO angle sharpens everything. Two million businesses already account for 40% of OpenAI's revenue. Getting that to 50% before listing is the goal. A superapp that keeps enterprise users inside one ecosystem is how you get there.

Our take:

The clearest way to understand what OpenAI is building is to look at Google. Think about how Google works. Gemini is embedded across Docs, Sheets, Slides, and Meet. Firebase and AI Studio handle developer-specific needs. NotebookLM got folded into the Gemini ecosystem. Everything Google builds lives inside one interconnected platform, and Gemini is the thread running through all of it.

That's exactly what OpenAI is attempting here. ChatGPT is the chat layer. Codex handles code. Atlas handles browsing. One company, one ecosystem, one model family powering all of it. The superapp is less a product launch and more an ecosystem declaration: we are done being a single chatbot, and we are building the Google Workspace of AI.

The open question is whether OpenAI can pull off what took Google two decades to build, before its IPO window closes and before Anthropic, Microsoft, and Google itself lock up the enterprise deals that make an ecosystem sticky in the first place.

FROM OUR PARTNERS

Most companies assume AI performance comes down to which platform they're running. Gladly analyzed top-performing retail AI deployments and found a different story. The biggest predictor of success wasn't the technology; it was what happened after launch. Download What retail's best AI deployments have in common to discover seven insights from real platform data.

Download the guide

🎓 AI Skill of the Day: Make Stunning AI Videos With Google Flow (Without Burning All Your Credits)

Google Flow is one of the most powerful AI video tools available right now, but most people open it once, feel overwhelmed, and close it. Paul J. Lipsky broke it down into a 15-minute tutorial that covers exactly what you need to get started fast.

The workflow is intentional: always start with images, then turn them into videos, then stitch clips into scenes. This order matters because image generation is cheaper on credits, so you nail the look first before spending on video.

Here's the core loop:

1. Create a new project and use the prompt box to generate images (pick your model, aspect ratio, and number of outputs)

2. Edit images by clicking into them and describing the change you want (e.g. "change the blue blanket to orange")

3. Use a prior image as a reference for new generations by clicking the + icon and attaching it to your prompt, which keeps character consistency across shots

4. Switch from image to video in the menu, attach your reference image, describe the action, and generate

5. Stitch clips together into a scene by clicking "add clip" and trimming start/end points on the timeline

6. Download scenes from within the scene view, not from the main media library

One credit-saving tip Lipsky flags: always check how many credits a generation will cost before hitting send. Pro plans get 1,000 credits/month; Ultra gets 10,000.

Total AI beginner? Start here (goes with this video).

Have a specific skill you want to learn? Request it here. 

🍪 Treats to Try

1. Ideogram 4.0 produces 2K output with dense text rendering across languages and precise object placement via bounding box control, so your generated images actually look designed, not AI-made, free tier available, paid plans from $8/month.

2. Boxes gives each Claude Code or Codex session its own isolated cloud computer so your agentic coding runs without conflicts, accessible from mobile or desktop, free tier available.

3. Google Labs added new tools including Flow (AI video creative studio), Stitch (vibe-to-UI design tool), and Pomelli (on-brand marketing content generator), all free to try while in early access.

4. Certified Fyn by OLO is a personal style curator for men that learns your taste and recommends outfits through a chat interface, free to try.

5. Empromptu lets teams build custom AI apps and fine-tuned models simultaneously, going from prompt to production-ready deployment in weeks with SOC 2 and HIPAA compliance baked in, pricing on request.

ICYMI: Vantor's Peter Wilczynski on the AI gap nobody's talking about.

Peter is CPO at Vantor, which has built a 100 million square kilometer 3D model of the entire Earth at 50cm resolution. The pitch: frontier AI models can write code and summarize documents, but they still can't understand the physical world. Vantor is building the bridge.

In this episode, Grant sits down with Peter to unpack spatial intelligence, how satellite imagery becomes machine-readable, GPS-free navigation with Vantor's Raptor system, and whether AGI actually needs to understand 3D space to be real AGI.

🎧 Watch / Listen: YouTube | Spotify | Apple Podcasts

📰 Around the Horn

• Banks are cutting junior analyst hiring classes by as much as two-thirds while simultaneously sourcing 62% of their new AI talent from those same cohorts, with Standard Chartered's CEO calling displaced workers "lower-value human capital" and Goldman's president describing operations as a "human assembly line ripe for automation."

• Microsoft launched seven in-house AI models called the MAI family at Build 2026, after renegotiating its OpenAI contract to remove restrictions that had blocked it from training its own frontier models.

• xAI secretly trained its coding models on Anthropic's Claude outputs for months, kept going through personal accounts after Anthropic cut off official access in January, and is now renting its GPU infrastructure to Anthropic and Google to pay the bills.

• Musicians union AFM sued Universal Music Group and Warner Music Group for licensing members' recordings to AI music generators Suno and Udio without compensating the artists, despite the labels pocketing settlement money from those same platforms.

• Intel CEO Lip-Bu Tan is betting Intel's revival on the rise of AI agents (software that completes multi-step tasks on your behalf), which is driving renewed demand for CPUs, Intel's core product, after years of losing ground to Nvidia and AMD.

FROM OUR PARTNERS

Flying blind: AI is failing because 71% of company workflows are invisible to leadership.

Scribe Optimize automatically detects real workflows using AI so you can spot inefficiencies and act on them. No surveys. No workshops. No guesswork. Most teams are optimizing processes they've never actually mapped.

Take a tour or request a demo — mention The Neuron to start your 30-day risk-free trial of Scribe Optimize.

😹Monday Meme

A Cat’s Commentary

P.S: Before you go… have you subscribed to our YouTube Channel? If not, can you?

P.P.S: Love the newsletter, but only want to get it once per week? Don’t unsubscribe—update your preferences here.