🕶️ F1 tech hit $2.4B. NVIDIA's in.

Platform owners are getting territorial. The speed at which companies are pulling critical layers back in-house is starting to look less like product strategy and more like existential reflex.

Today: a full-stack rewrite of search, simulation, and staffing, plus a wave of exits from dependencies that suddenly felt expensive. What looked like partnership six months ago is turning into platform consolidation, and if you're sitting in the wrapper economy, the ground just shifted.

-🕶️

Right now, pay just $20 line when you switch 10 or more lines on T-Mobile for Business CoreMobile plan.

Bring your device and keep your current numbers.
Plus, get an additional $150/line back when you switch. No limits, so you do the math.

*10 lines for $200/mo; each additional line $20/mo./line with Autopay; plus taxes & $35 device connection charge. $150/line credit via one-time bill credit.

PhysicsX just raised a $300M Series C, but the real story is simpler: if you can turn simulations that took days into seconds, you are not selling “AI”; you are selling faster engineering.

Temasek led at a $2.4B valuation, with Nvidia, Applied Materials, and Siemens increasing their stakes. That matters more than the headline number. Strategic investors usually re-up when the product is already inside a critical workflow, not when they are chasing a trend. Two years after its $32M Series A, PhysicsX says headcount doubled to 350 and revenue quadrupled.

The most useful signal: it is supply-constrained. Jacomo Corbo says rollout is being moderated because demand is ahead of delivery capacity. That is a very different problem from weak adoption. Semiconductors becoming the largest vertical this year tells you where this is going.

The downstream implication is hard to miss. The companies building chips, turbines, cooling, and data center infrastructure now need an AI-native simulation layer to move faster. If your customer’s bottleneck is physics, and you sell compressed design cycles, the moat is not branding. It is the math.

Apple is doing what platform owners always do once a layer gets strategic: pulling it in-house.

At WWDC 2026, Siri is expected to become a standalone app across iOS 27, iPadOS 27, and macOS 27, with Apple answering queries itself instead of routing queries externally. That means synthesized answers, summaries, images, and on-device retrieval without Google in the loop and without a ChatGPT handoff.

The real signal is the speed; nine months ago, ChatGPT integration was Apple’s headline AI move; now Apple appears to be replacing the partner layer with its own query-to-answer stack. That is a very fast shift from dependency to control.

If you are building retrieval-augmented search, this is the warning shot. iOS owns 57% of the U.S. smartphone market. If Siri stops exporting intent, outside answer engines lose a major distribution path. Google’s $20B+ default-search economics also get tighter, no matter what happens in court.

For founders, the takeaway is simple: if your product is just a wrapper on someone else’s search or model API, the floor is dropping. The safer ground is proprietary data, domain-specific workflows, or the enterprise integration layer Apple will not cover well.

Scale your growth: Join names like HubSpot, Ahrefs, and Deel. Put your product in front of 90,000+ startup founders. Let's talk.