2Instagram | Sponsor this newsletter | Let’s Connect!

Meta dropped nearly $15 billion on a strategic bet that could change the entire AI ecosystem. Apple finally gave us a deeper look at how “Apple Intelligence” will quietly take over your device. And MidJourney? Yep—they’re getting into video now.

From the kitchen to the cloud, AI is moving fast—and getting weird. Let’s unpack the biggest breakthroughs, smartest tools, and the moments that made us say: “Did my fridge just talk to me?”

Apple’s WWDC reveals on-device AI that’s private, practical, and… surprisingly good.

Apple’s new features didn’t come with splashy AGI demos—but their AI tools are quietly becoming indispensable. And they all run on-device.

Apple’s positioning itself as the private AI company—running powerful models directly on your iPhone. Even the Apple Watch gets smarter workouts, and Apple Vision Pro now has persistent AI widgets and spatial memory.

🔮 Prediction: Apple’s long-term play is clear—AI that feels invisible but deeply personalized. Expect more devs to flock to its on-device models, building tools that feel native and frictionless.

👉 Read Apple’s full AI update

MidJourney video is finally real—sort of.
For months, MidJourney teased generative video. Now, they’ve launched a video ranking party—where users rate model outputs to fine-tune quality.

Meanwhile, Meta AI launched preset-based video editing—transforming outfits, lighting, and environments in real footage. Think Gen-1 and Runway, but baked into Facebook’s core.

🔮 Prediction: Expect MidJourney to launch text-to-video within weeks. Meta’s going wide with consumer access, while MidJourney is going deep on quality. The generative video wars are just getting started.

👉 Try MidJourney’s video ranking
👉 Edit your video with Meta AI

Meta acquires 49% of Scale AI. This isn’t just an investment—it’s a land grab.
Scale AI provides training data to nearly every major AI company—including OpenAI, Microsoft, and Nvidia. Now Meta owns nearly half of it.

Scale says it’ll stay “independent”—but when your biggest competitor owns half the shop, things get awkward fast.

🔮 Prediction: Meta’s planning a vertically integrated AI empire, from infrastructure to intelligence. This is their OpenAI–Microsoft moment—and it may force other labs to rethink who trains their models.

👉 See Scale’s official announcement

Want AI to actually do something with your time?

Here’s how to make Google Gemini auto-schedule your creative or work tasks.

💡 Why it matters: This moves AI from chat toy to true assistant. Instead of reactively asking questions, Gemini starts acting proactively—turning static prompts into workflows.

👉 Try it in the Gemini App

🧠 OpenAI Pro users just got access to the O3 Pro model—its fastest, most capable reasoning model yet. Read update
🌩️ DeepMind’s Weather Lab now forecasts cyclones 15 days out—across 50 scenarios. Explore more
📸 Leonardo AI’s new Lucid Realism model rivals photo quality—with eerily lifelike image generation. See samples
🧠 Mistral's Magistral is Europe’s first reasoning LLM—with open weights and enterprise variants. Test here
🎮 Disney and Universal sue MidJourney for AI-generated IP violations—highlighting a murky future for copyright law. Full story

Until our next AI rendezvous,

Anthony | Founder of Uncover AI