WhoKnows.
← All briefings
MONEYTECH 3 stories

Daily Briefing — March 25, 2026


01

Arm Is Now Making Its Own Chips

Wired →
Tech shifts + Money & markets

Arm just blew up its own business model. For decades, the company made money by licensing its chip designs to others and letting them do the actual building and selling. Clean, asset light, very profitable. Now Arm CEO Rene Haas stood in front of a live audience in San Francisco and held up a physical chip, saying out loud that Arm is now in the business of supplying CPUs directly. That is a significant line to cross.

The new chip is called the Arm AGI CPU, which tells you everything about where they think the puck is going. It is designed for data centers, meant to handle agentic AI workloads, and is being fabricated by TSMC. So Arm is not just dipping a toe in. They went straight to the world's best chip manufacturer and built something meant to compete in the most contested computing market on the planet right now.

Here is the part worth sitting with. Arm's existing customers, the Qualcomms and Apples of the world, license Arm designs to build their own chips. Now Arm is becoming a competitor to those same customers. That tension does not resolve quietly. Companies across the AI infrastructure stack are going to have to rethink their supplier relationships, their build versus buy calculations, and frankly who they trust.

SO WHAT

If you work anywhere near AI infrastructure, cloud platforms, or enterprise hardware decisions, the vendor landscape you mapped out last year just started shifting underneath you.

ACTION ITEM

This week, look up how your company or your clients currently source compute for AI workloads and flag whether any of those decisions assume Arm stays a pure licensing business, because that assumption just expired.


02

SpaceX Could File For Mammoth IPO This Week: The Information

Investor's Business Daily →
Money & markets + Tech shifts

SpaceX might be about to do something that makes every other tech IPO look like a lemonade stand. According to The Information, Elon Musk's rocket company could file for an IPO as early as this week, with a target raise of $75 billion. That would make it one of the largest public offerings in history, full stop.

Here is the thing though. SpaceX is not just a rocket company anymore. It runs Starlink, which is a global satellite internet business with real revenue and real subscribers. It has government contracts. It is building infrastructure that countries and militaries depend on. This is not a vibe based startup with a deck and a dream. So when a company like this goes public, it reshapes how investors, employers, and entire industries think about what "big tech" even means.

For people working in FinTech, finance, or any adjacent space, a listing of this scale sends ripples. It signals that the market has an appetite for long horizon, capital intensive bets again. It also means a wave of attention, talent, and money is about to flow toward the aerospace and deep tech sectors in ways that will affect hiring, valuations, and where the smart money thinks the next decade is being built.

SO WHAT

If you work in tech, finance, or product, this IPO is going to reframe the conversation about which industries are worth betting your career on for the next decade.

ACTION ITEM

This week, spend 30 minutes reading up on SpaceX's Starlink revenue model so you can speak intelligently about it when it inevitably comes up in your next meeting or interview.


03

OpenAI announces plans to shut down its Sora video generator

Ars Technica →
Tech shifts + Money & markets

OpenAI is shutting down Sora, the text to video generator it launched to massive fanfare in late 2024. The company confirmed it on social media right after the Wall Street Journal broke the story, with the standard "thank you to our community" language that companies always reach for when they're killing something people actually liked.

Here's the part worth paying attention to though. This isn't just a product being discontinued. It's a signal about where OpenAI's head is at right now. Executives reportedly told staff at a recent all hands meeting that the company needs to stop being "distracted by side quests" and refocus on business and productivity tools. Sora, for all its wow factor, was basically a very expensive demo that hadn't found a clear enterprise use case yet.

The Disney angle makes this even messier. Disney dropped a billion dollars into OpenAI just months ago, and part of that deal explicitly involved bringing Disney characters to life through Sora. Nobody has explained yet what happens to that partnership now. That's a very awkward conversation happening somewhere in a very expensive conference room right now.

The broader lesson here is that in AI right now, the gap between "impressive launch" and "sustainable product" is enormous. Tools you build workflows around can disappear faster than you think.

_Forwarded this? Subscribe at [whoknows.news](https://whoknows.news)_

_You're receiving this because you subscribed. [Unsubscribe]({{unsubscribe_url}})_

SO WHAT

If you or your team have been building creative workflows or client deliverables around any single AI tool, Sora's shutdown is a reminder that your process is only as stable as the platform underneath it.

ACTION ITEM

This week, audit one AI tool your work actually depends on and identify a backup option so you are never one shutdown notice away from a broken workflow.