I spend a ridiculous amount of time reading through AI whitepapers, release notes, and tech leaks, but this week legitimately made me stop and re-evaluate everything I thought I knew about the current trajectory of the industry. We aren’t just talking about fun new chatbots anymore; we are watching the foundation of the global economy, military infrastructure, and the job market shift in real-time.
Let me walk you through the most mind-bending developments of the week, skip the corporate fluff, and tell you what this actually means for us.
The Pentagon Drama: A Divide in the AI World
The biggest shockwave this week wasn’t a new model; it was a massive philosophical split. OpenAI’s CEO, Sam Altman, struck an agreement with the Trump administration to integrate OpenAI’s models into the Pentagon’s classified networks for “any legal purpose.”
Honestly, I’m not entirely surprised by the move, but the backlash was immediate and fierce. The “QuitGPT” movement exploded across X and Reddit, with thousands of users canceling their subscriptions in protest.
But here is the fascinating part: Anthropic capitalized on this perfectly. By publicly refusing to give the Pentagon unlimited control, they positioned themselves as the ethical alternative. As a result, users flooded to Claude, pushing the app all the way to the #2 spot on the App Store this week. It’s a clear signal that the AI race isn’t just about parameter counts anymore—it’s about trust.
Perplexity Computer: 19 Brains in One
If you want to understand where AI agents are heading, look at what Perplexity just dropped. They introduced “Perplexity Computer,” an absolute beast of a system available to their Max subscribers (for a hefty $200/month).
Instead of relying on a single AI model to do everything passably well, this system orchestrates 19 different AI models simultaneously. It delegates coding to the best coding model, research to the best search model, and design to the best visual model. It has its own file system, command-line interface, and real-time browser access.
In my view, this isn’t just an AI assistant anymore; it’s literally a cloud computer running on neural networks. It can research, design, code, deploy, and manage a project from start to finish. If you’ve been waiting for true AI agents, they are officially here.
Jack Dorsey’s Brutal AI Reality Check
I always try to stay optimistic about technology creating new opportunities, but we have to face reality: AI is replacing jobs right now.
Jack Dorsey’s company, Block, just announced the most aggressive downsizing in its history, cutting its global workforce from over 10,000 to under 6,000. That’s more than 4,000 people. Dorsey didn’t sugarcoat it either; he explicitly stated that AI fundamentally changes what it means to build and run a company.
The grim part? Wall Street loved it. Block’s stock surged up to 24% after the announcement. This tells me that investors are actively rewarding companies that replace human labor with AI efficiency.
The “Ghost GDP” Threat
Speaking of the economy, the most terrifying thing I read this week came from Citrini Research. They published a thought experiment called The 2028 Global Intelligence Crisis, and it kept me up at night.
They introduced a concept called “Ghost GDP.”
Here is how it works: By 2026, white-collar layoffs push unemployment past 10%. On paper, national productivity and GDP look incredible because AI agents are generating massive amounts of output. But this wealth is “ghost” wealth—it never enters the real economy. Production skyrockets, but because humans aren’t earning wages for that production, they aren’t spending money. The output doesn’t trickle down through traditional income channels, creating a massive, hollow bubble.
Even though it’s just a theoretical scenario, it caused actual ripples in the market this week. It made me realize that our current economic systems are completely unprepared for agent-driven production.
Hardware Leaps: Powering the Next Generation
Of course, none of this software magic happens without silicon. Here’s what’s happening on the physical front:
- Nvidia’s Vera Rubin: Nvidia has officially started shipping early samples of its next-gen Vera Rubin AI platform to select clients. Packing an 88-core CPU and Rubin GPUs with up to 288GB of HBM4 memory, this is the architecture that will power the AI data centers of 2026 and 2027.
- Snapdragon Wear Elite: Qualcomm is bringing heavy-duty AI to our wrists. The new Wear Elite chip features a dedicated built-in NPU capable of running AI models with up to two billion parameters right on your smartwatch. We’re talking instant text generation, smart summaries, and real-time coaching without needing to ping a cloud server.
Quick Hits: What Else I Tracked This Week
The news cycle was so fast I barely had time to process it all. Here are the other massive updates you need to know:
- The $26 Billion Claude Code Shock: IBM’s stock crashed, wiping out $26 billion in market value. Why? Because Anthropic’s Claude Code just proved it can effortlessly read and modernize ancient Cobol programming—a massively lucrative legacy service IBM used to dominate.
- Google’s Visual Leap: Google launched Nano Banana 2, which is hands down their most advanced visual generation tool to date. The fidelity is insane.
- Massive AI Investments: OpenAI is wrapping up one of the biggest funding rounds in Silicon Valley history, locking in a staggering $110 billion commitment from Amazon, NVIDIA, and SoftBank.
- ChatGPT’s User Base: OpenAI casually mentioned they hit 900 million weekly active users, with over 5% paying for subscriptions. They are also reportedly working on a middle-tier “Pro Lite” plan.
- DeepSeek V4 is Coming: Financial Times leaked that DeepSeek is launching its multimodal V4 model in early March. If it’s anything like R1, the industry is in for another pricing shock.
- Green AI Servers: Google announced their new Texas data center will use advanced air-cooling tech to drastically cut down on water consumption—a much-needed win for the environment.
When I step back and look at this week’s timeline—from the Pentagon contracts to the Perplexity cloud computer and the chilling “Ghost GDP” theory—it feels like we just crossed a point of no return. The infrastructure of the future is being laid down right beneath our feet.
I’m incredibly curious about your take on the Citrini Research report. Do you think “Ghost GDP” is a genuine threat to our economy, or will humans naturally find new, currently unimaginable ways to generate value alongside AI? Let me know in the comments.
You Might Also Like;
