Meta is giving its discontinued Portal devices a second life. And turning one of its forgotten hardware products into a testing ground for the agent era.
Most companies kill failed hardware and move on. Meta is doing something more interesting.
The company has announced new AI-powered developer tools that allow builders to repurpose old Portal devices into smart home dashboards, family message boards, AI assistants, and other custom applications. The move effectively transforms a discontinued product into an experimental platform for agentic AI.
At first glance, this looks like a clever way to recycle unused hardware. But it’s not.
The more important story is what Meta appears to be learning from the AI race.
For years, the company approached hardware as a destination. Portal was supposed to be a consumer product. Consumers never really bought into the vision. Privacy concerns followed the device from launch, and Meta eventually discontinued the product as it shifted its focus elsewhere.
But AI is changing the economics of hardware.
Suddenly, a screen, a camera, microphones, and an internet connection are enough to create something useful. The value no longer comes from the device itself. It comes from the intelligence running on top of it.
That’s why this announcement feels larger than Portal.
Across the industry, companies are trying to figure out what AI agents actually need to exist in the physical world. Not every interaction belongs on a laptop. Not every request should happen through a smartphone. Sometimes the ideal interface is simply a screen in the kitchen, office, or living room that’s always available and context-aware.
Meta seems to be experimenting with exactly that idea.
What’s notable is that the company says these tools are hardware-agnostic. That suggests Portal may be less of a product revival and more of a proving ground for future devices. The company can learn how people use AI assistants in physical spaces without building entirely new hardware from scratch.
For tech buyers, the announcement points toward a broader shift that’s beginning to emerge across enterprise and consumer technology alike.
The conversation around AI has largely focused on models. Which model is smartest? Which one reasons better? Which one generates better outputs?
The next phase may focus on surfaces.
Where does AI live? Which devices become the primary interface? How many existing endpoints can be turned into AI-native experiences instead of being replaced altogether?
That matters because organizations are sitting on thousands of screens, kiosks, tablets, conference room displays, and edge devices. If AI can extend the life of existing hardware, the economics of AI deployment start to look very different.
Instead of asking what new hardware they need to buy, technology leaders may begin asking what hardware they already own.
Portal’s second life hints at a future where AI doesn’t just create new products. It gives old ones a reason to exist again.
