What if your computer stopped waiting for you to click things and just… did stuff on your behalf?
That's the pitch behind Nvidia's latest move into consumer hardware. And while the tech world is having its "iPhone moment" celebrations, I think designers should be paying close attention — not because it's exciting, but because the assumptions baked into this vision reveal a lot about how the industry thinks about creative work.
What Happened
Nvidia just unveiled the RTX Spark, a chip that combines a microprocessor and graphics chip into a single package designed to run AI agents locally on your laptop. No cloud dependency. No sending your work to someone else's server. Built in partnership with Microsoft and MediaTek, it'll ship this year inside machines from Dell, Lenovo, Asus, and HP.
The big sell? CEO Jensen Huang called it a reimagining of the PC "for the first time in 40 years." The idea is that instead of you opening apps, navigating menus, and clicking buttons, AI agents will handle tasks autonomously. One analyst compared the moment to the arrival of the iPhone.
That's quite a claim.
The Bigger Picture
Here's what I find interesting — and slightly troubling — about the "agentic PC" framing.
The entire pitch assumes that the bottleneck in computing is you. That clicking and typing are the problem, and removing human interaction from the loop is the solution. For spreadsheet jockeys and email wranglers, maybe that's partly true. But for designers?
We're spatial thinkers. We manipulate things directly. We nudge a layout two pixels to the left because it feels better. We pick a colour not because an algorithm says it's optimal but because it carries a specific emotional weight. The mouse and keyboard aren't obstacles to our work — they're instruments of it.
So when someone says "agents will replace traditional mouse and keyboard interactions," I'd encourage you to ask: replace them with what, exactly?
That said, I don't think this is all hype. The genuinely interesting part isn't the agent fantasy — it's the local processing.
Right now, most AI-powered design tools lean heavily on cloud infrastructure. Every time you generate an image, run a style transfer, or ask an AI to iterate on a layout, your data takes a round trip to someone else's data centre. That means latency, privacy concerns, and a hard dependency on your internet connection.
Local AI changes that equation. Imagine running Stable Diffusion-level image generation natively on your laptop without it sounding like a jet engine. Or having real-time AI-assisted video editing that doesn't buffer. Or — and this is the one I'm most interested in — keeping sensitive client work entirely on your machine while still using AI tools.
That's not a revolution in how we design. It's a revolution in where the computation happens. And for freelancers and studios handling confidential work, that distinction matters enormously.