🌍 Google just made reality promptable

Google had a massive week, but the bigger story is not just one launch.

AI is starting to become the layer between imagination and output: videos, apps, simulated cities, enterprise workflows, and even the infrastructure bill behind all of it.

Here are the stories worth knowing.

Google introduced Gemini Omni at I/O, a new “any-to-any” world model that can generate video, audio, and text from mixed inputs. You can feed it a prompt, reference images, voice notes, or audio clips, and it can create a finished video with matching sound, music, voice-over, and stable on-screen text.

The biggest upgrade is control. Omni supports conversational editing, so users can keep refining a scene without starting from scratch. Add an object, change the background, swap a style, or tweak the camera while keeping the broader scene intact.

This is Google making a serious run at AI video.

For creators and brands, this compresses multiple tools into one creative system. For everyone else, it makes believable AI video much easier to produce.

Google also expanded Project Genie by connecting it to nearly 20 years of Street View imagery. Users can drop a pin on a real map, apply a style, describe a character, and explore a generated 3D simulation based on the actual streets and buildings.

The system uses Google’s massive map archive to preserve real-world geometry while changing the surface layer. A normal city block can become a desert wasteland, cloud city, underwater world, or game-like environment.

This looks like entertainment, but the uses go much deeper.

Google is turning maps into simulation fuel.

Google AI Studio now lets users build native Android apps from a natural language prompt. It generates real Kotlin and Jetpack Compose code, previews the result in a browser-based emulator, and can even deploy to a connected phone.

The system is not just making mockups. It can build apps that access real phone hardware like GPS, Bluetooth, and NFC. If the code fails, an AI agent loops through errors and tries to fix the build automatically.

This is one of the clearer signs that app development is becoming more accessible.

The interesting part is not that everyone becomes a perfect app developer overnight. It is that more people can get to a working prototype without waiting for a team.

A SpaceX filing reportedly revealed that Anthropic is paying $1.25 billion per month for AI compute through a major infrastructure deal. The contract runs through May 2029 and gives Anthropic access to hundreds of thousands of advanced Nvidia GPUs and hundreds of megawatts of power.

The unusual twist: the capacity is tied to infrastructure originally built around Elon Musk’s xAI ambitions, with SpaceX now acting more like a high-end AI cloud landlord.

The AI product you see in the app is just the front end. Behind it is a monthly bill big enough to fund entire companies.

Meta has begun laying off around 8,000 employees, roughly 10% of its workforce, as part of a restructuring built around AI. The company is also canceling thousands of open roles and shifting remaining employees into higher-priority AI divisions.

The cuts come as Meta plans enormous AI infrastructure spending, including data centers, GPU clusters, and long-term superintelligence projects.

This is the human cost of the AI buildout.

Meta is not shrinking because it has no money. It is choosing where the next decade’s money goes.

AI is becoming the operating layer.

It can turn prompts into videos, maps into simulations, ideas into apps, and enterprise demand into billion-dollar monthly compute contracts. But the cost is becoming harder to ignore: more infrastructure, more power, and fewer people in certain parts of Big Tech.

The next phase will not be defined only by better models. It will be defined by who can turn those models into systems people actually use, and who can afford the machine behind them.

If this issue helped you make sense of AI’s chaos, forward it to a friend who shouldn’t be sleeping on this.

Until next time,
Long Live AI