Apple's AI, Grok Goes Open

· The Fluency Briefing

Welcome back to your weekly dose

This Week in AI

Hey there — this was the week AI stopped being a sector and started acting like the economy. Forbes dropped a bombshell stat about AI companies rewriting the corporate scoreboard, OpenAI revealed a custom chip called Jalapeno, and Oracle quietly confirmed 21,000 job cuts driven by AI adoption. Meanwhile, the U.S. government yanked two Anthropic models off the market. Chaos, ambition, and consequences — all in seven days. Let's break it down.

Weekly Theme

📰 The Big Story

Here's the thing about AI's economic moment: it's no longer just hype reflected in stock tickers. Forbes reported that AI-linked companies drove over half of all new market value growth across the Global 2000 — the annual ranking of the world's largest public companies forbes.com, Jun 24. Let that land. More than half of the planet's corporate value creation now traces back to AI. That's not a trend; that's a restructuring.

And the companies riding this wave aren't content to rent their future. OpenAI, partnering with Broadcom, unveiled "Jalapeno," a custom inference chip built from the ground up for large language models, with early testing showing substantially better performance per watt than current hardware openai.com, Jun 24. Days later, Qualcomm dropped nearly $4 billion to acquire chip startup Modular wired.com, Jun 24, while IBM showcased a sub-1nm chip design cramming nearly 100 billion transistors onto a fingernail-sized surface bbc.co.uk, Jun 25.

First order: AI companies want to own their silicon. Second order: this vertical integration squeezes out smaller players who rely on commodity chips. Third order: the firms that control both the models and the metal underneath them will set the pricing, the pace, and the rules for everyone else.

For you, this means the cost and capability of the AI tools you use in 12 months will be shaped less by software breakthroughs and more by who won the chip war happening right now. The companies building their own hardware stacks aren't just optimizing — they're building moats.

Reaction

📋 5 Stories That Shaped the Week

Beyond the headlines, here's what shaped the week...

Oracle disclosed it cut 21,000 employees over the past year — a 13% reduction — and explicitly blamed AI adoption for the losses techcrunch.com, Jun 23. Let's be real: this isn't a one-off. It's a case study in what happens when a legacy enterprise decides AI can do what people used to. The "so what" here isn't about Oracle. It's that every company running similar workloads is doing the same math right now.

While Oracle was trimming, the U.S. government was pulling the emergency brake on Anthropic. Federal authorities forced the company to yank its Claude Fable 5 and Mythos 5 models after Amazon researchers allegedly found jailbreak vulnerabilities techcrunch.com, Jun 20. But here's the wrinkle — as TechCrunch noted, export controls on AI have a long, futile history techcrunch.com, Jun 20. Translation: the ban made a political statement, but whether it actually stops anything is another question entirely.

Apple, meanwhile, unveiled a rebuilt Siri AI at WWDC 2026. Early hands-on reviews describe a genuinely useful assistant with deep cross-device and app integration wired.com, Jun 20. Apple also launched several companion apps designed to embed intelligence across its ecosystem macrumors.com, Jun 21. The real story? Apple is betting that AI's killer app isn't a chatbot — it's invisible help woven into the devices you already own.

Across the Atlantic, Europe is doubling down on domestic AI infrastructure under the banner of tech sovereignty sifted.eu, Jun 22. And the buildout isn't just a policy aspiration — it's hitting real-world bottlenecks. Data center construction faces critical skilled labor shortages even as billions pour in tomshardware.com, Jun 24. Nvidia CEO Jensen Huang argued that electricians and plumbers will be needed "by the hundreds of thousands" in this new economy fortune.com, Jun 20, a reminder that AI's physical layer still depends on very human hands.

🔗 The Pattern We Noticed

Connecting the dots...

The thread running through this week is vertical integration at all costs. OpenAI is designing its own chips. Apple is rebuilding Siri to own the interface layer. Europe is building sovereign infrastructure to avoid dependence on U.S. hyperscalers. Even Oracle's layoffs reflect the same impulse — collapsing the distance between capability and control.

Why now? Because the companies that rent their AI stack — the chips, the models, the distribution — are realizing they've handed their margins and their strategic flexibility to someone else. In the gold rush phase, that was fine. In the profitability phase, it's fatal.

For you, this means vendor lock-in is the quiet risk of 2026. The tools you adopt today come with assumptions about whose chips they run on, whose cloud they live in, and whose rules they follow. Before you double down on any AI platform, ask: who actually controls the stack underneath it? That question will matter more than any feature comparison.

Meme

🔮 On the Horizon

These stories are still unfolding — here's what to track:

📚 Term of the Week

Term illustration

Going deeper on one concept that shaped this week's AI conversation.

"Vertical Integration"

What it is: Vertical integration means a company controls multiple stages of its own supply chain — from raw materials or components all the way to the end product. In AI, this looks like a company designing its own chips, building its own data centers, training its own models, and distributing through its own apps, rather than relying on third-party providers at each step.

Why it matters this week: OpenAI building custom chips, Qualcomm acquiring Modular, and Europe pursuing sovereign infrastructure all signal the same play — owning the stack from silicon to software.

The bigger picture: As AI shifts from experimentation to core business infrastructure, vertical integration determines who captures value and who becomes a commodity supplier. The companies that control their full stack will set pricing, performance benchmarks, and access rules for everyone else.

Try this: Map out one AI tool you use. Ask: who made the model, who hosts it, who designed the chip? Count how many companies you actually depend on.

📬 That's a Wrap

That's a wrap on this week — a week where AI flexed economic muscle, built its own chips, and showed us that growth and disruption are two sides of the same silicon wafer. The question isn't whether AI is transforming the economy. It's whether you're positioned on the right side of the transformation.

Your move: Pick one AI tool your business relies on and trace its supply chain — model provider, cloud host, chip maker. If one company controls all three layers, that's a concentration risk worth understanding now.

Fluently yours, The My AI Fluency Team

What We're Working On

✨ Founding Cohort Special - 60% Off! — Use code MAF20 to join for just $20/month (regularly $50). Get weekly group sessions & workshops, self-paced courses for all levels, access to tools & templates, challenges with peer feedback, and 24/7 support community. → Join Now

✨ Free 30-Minute AI Consultation — Discover how My AI Fluency can help your business unlock the potential of AI. We'll discuss your goals, explore practical AI opportunities for your industry, and outline clear next steps. → Schedule Free Call

✨ How AI-Fluent Are You? — Test your AI fluency with our interactive quiz. See how you stack up and discover what to learn next. → Take the Quiz

💬 Community | 📞 Book a Consultation | 🌐 Website

My AI Fluency