Jack Dorsey just mass-fired half his company

Block—the company behind Square and Cash App—just laid off 4,000 people. That's nearly half the entire workforce. Gone. And Jack Dorsey didn't blame "macroeconomic headwinds" or "strategic restructuring." He blamed AI.

His exact words: "Intelligence tools have changed what it means to build and run a company."

He said in December the models "got an order of magnitude more capable and more intelligent," and he wanted to act decisively rather than "cut gradually over months or years as this shift plays out." Translation: rip the band-aid off now instead of bleeding out slowly.

The prediction that should scare you: Dorsey said within the next year, the majority of companies will reach the same conclusion and make similar structural changes. Not some companies. Most companies. Block went from over 10,000 employees to just under 6,000 in a single announcement. The stock surged. Wall Street sent flowers.

The counter-argument is worth reading: CNN is calling this potential "AI-washing"—dressing up old-fashioned cost-cutting as technological futurism. Block's revenue growth had been slowing. Their Cash App lending business was under pressure. Dorsey had already been criticized for being spread too thin across companies. Did AI really just "change what it means to build a company"—or did it give a CEO a narrative that makes Wall Street clap instead of asking uncomfortable questions?

The honest answer is probably both. AI is a real force multiplier. It's also a convenient story. And the uncomfortable part is that it doesn't matter which one is true for the 4,000 people who just lost their jobs.

The pattern is already here. Block isn't an outlier. It's the latest entry in a growing list. Klarna cut its workforce and bragged about replacing support agents with AI. Duolingo dropped contractors for AI-generated content. Salesforce froze engineering hiring entirely. Goldman Sachs is testing agentic AI for trade surveillance. A Joint Venture Silicon Valley report flagged 410,000+ jobs in the valley alone that are exposed to AI-replaceable tasks.

And it's not just tech. The Atlanta Fed's Raphael Bostic has started publicly discussing AI's impact on structural unemployment. Challenger, Gray & Christmas is now tracking AI as a standalone category in layoff data. This isn't a blip. It's a category.

The question that actually matters: If Dorsey's right—if most companies reach the same conclusion within a year—what does "making it" look like? It's not about being irreplaceable. Nothing is irreplaceable. It's about being the person who can do the work of three because you figured out how to build with AI, not just chat with it. The gap between "I use ChatGPT sometimes" and "I ship things with AI daily" is where the next round of cuts will land.

That's not fear-mongering. That's the lesson from every company on the list above. The people who survived weren't the most senior or the most experienced. They were the most AI-fluent.

So what do you actually do about it? You become AI-native. Not tomorrow, not Q3, not "when things settle down." Now. We break down exactly what that means—and how to get there—below. 👇

Everyone's throwing "AI-native" around like it means something obvious. It doesn't. So let's make it concrete.

The simplest definition: AI-native means AI is your default, not your add-on. The same way you don't "use the internet"—you just work on it—AI-native means AI is baked into how you think, build, and ship. It's not a tool you open. It's the operating layer underneath everything.

The analogy that clicks: Remember when companies had a "website guy"? One person who knew HTML and maintained the company site? Then the internet became the operating layer for everything, and suddenly everyone needed to be internet-literate. The companies that said "we have a web team, we're covered" got lapped by the ones where everyone lived on the internet natively. AI is at that exact inflection point right now. The companies with an "AI guy" are about to get eaten by the ones where everyone is AI-native.

Here's the actual framework. Three levels. Be honest about where you are.

You talk to Claude or ChatGPT like a thought partner. You draft with it, brainstorm with it, pressure-test ideas with it. You paste in a doc and say "what am I missing." You ask it to poke holes in your strategy before your boss does. This is table stakes in 2026. If you're not here yet, you're behind.

How to get here today: Open Claude or ChatGPT. Paste in the last thing you wrote—email, strategy doc, project brief—and ask: "What's the weakest argument in this? What would a skeptic say?" Do this once a day for a week. You'll never write without it again.

You don't just chat—you ship. Landing pages, competitive briefs, data analysis, presentations, automated workflows. You're using tools like Claude Code, Cursor, or Cowork to create real deliverables that used to require a team or a contractor. This is the level where you go from "person who uses AI" to "person who does the work of three." This is where the Dorsey cuts hit. The survivors are here.

How to get here this week: Pick one task you currently outsource or delegate—a report, a landing page, a competitive analysis—and build it yourself with AI. Our Claude Code 101 playbook walks you from zero terminal experience to running five AI agents in parallel. If you don't want to touch a terminal, How to Ship Software Without Touching Your Keyboard shows you Ryan Carson's 2-file system for turning LLMs into junior engineers.

You're building agents and workflows that run while you sleep. Customer health monitors that flag churn risk before your CSMs notice. Lead enrichment pipelines that score and route without a human. Content workflows that draft, edit, and schedule across channels. This is where it gets unfair—where one person genuinely replaces a team, not as a talking point but as a P&L line item.

How to get here this month: Start with one repeatable workflow. Something you do every week that follows a predictable pattern. Build an agent for it. Our One-Person Growth Team playbook shows you how to build bulk ad generation, landing page pipelines, and lead enrichment as a one-person team. If you want the strategic framework, Build an AI Strategy That Doesn't Suck is the 6-step playbook from executive buy-in to shipping.

The uncomfortable truth: Most people are stuck between levels 1 and 2. They chat with AI but haven't made the jump to building with it. That gap—between "I prompt sometimes" and "I ship with AI daily"—is exactly where the layoffs land. It's also the easiest gap to close. You don't need to learn to code. You don't need a CS degree. You need a weekend and the willingness to feel stupid for a few hours.

Pick your lane and go:

OpenAI raised $110B—the largest private round in history — $730B valuation. Amazon put in $50B, Nvidia and SoftBank added $30B each. But here's the clause: $35B of Amazon's commitment only kicks in if OpenAI goes public—or builds AGI. Amazon is literally betting $35 billion on the possibility of artificial general intelligence. Meanwhile, OpenAI signed a Pentagon deal the same day Anthropic got axed—with basically the same safety red lines Anthropic got banned for having. ChatGPT is at 900M weekly active users. 50M paying. TechCrunch

Computer-use agents are suddenly everywhere — Cursor Agents now test their own work and send you a video demo. Cursor acquired Autotab, Anthropic acquired Vercept—both for computer-using agents. Perplexity launched "Perplexity Computer." Google previewed Gemini ordering food autonomously on a Galaxy S26. The pattern: agents are moving from CLI to GUI, from typing commands to clicking buttons. This is how AI goes from "tool for engineers" to "tool for everyone." The New Stack's roundup

Google dropped Nano Banana 2 — Google's new default image generation model combines Nano Banana Pro's quality with Gemini Flash speed. Text rendering in images is finally legible (95%+ accuracy), it pulls from real-time web knowledge, and it's rolling out across Gemini, Search, and Ads. If you're generating marketing mockups or ad creatives with AI, this is the new baseline.

Claude Code got remote control — Start a session in your terminal, pick it up later from your phone or the web app. It keeps working on the original machine. If you're running long tasks, you can finally walk away from your desk without killing the process. Docs

Suno hit $300M ARR — The AI music generator now has 2 million paid subscribers. Creative AI tools are quietly printing money while everyone debates whether AI can "really" be creative. TechCrunch

Supreme Court won't touch AI copyright — Declined to hear whether AI-generated art can be copyrighted. Current rules stand: human authorship required. If you're building with generative AI, this matters for everything from marketing assets to product design. Constitution Center