You know that feeling when you pick up your phone to check one thing and somehow end up watching videos for 45 minutes? Yeah. Well, OpenAI, the company behind ChatGPT is apparently trying to build something that fixes exactly that. Or at least, that's the idea.
So basically, OpenAI is building a completely new device. Not a laptop. Not exactly a smartphone either. The short version is: they want to create a gadget where you talk to AI the way you'd talk to a person, and it just... handles things. No apps, no swiping, no screen full of notifications. The longer version is a bit more of a mess, and honestly the whole thing is still being figured out.
Let me break it down properly on what OpenAI is actually making.
Wait, Who Is Making This and Why Should I Care?
OpenAI is the company that made ChatGPT, which you've probably used. Last year in May 2025, they spent $6.5 billion to buy a small hardware startup called "io", which was founded by a designer named Jony Ive.
Now, Jony Ive is kind of a big deal in the tech world. He's the person who designed the original iPhone, the iPod, the iMac, basically all the Apple products that people actually liked.
The idea is that Ive brings the design knowledge and OpenAI brings the AI brains. Sam Altman (OpenAI's CEO) said in November 2025 that the early prototypes were "jaw-dropping good."
What they're working on is not one device. It's actually a few different things. There's a small audio companion device, the size of an iPod Shuffle, so really small, you can put it in your pocket or wear it around your neck. There's apparently also an AI pen, which is codenamed "Gumdrop" internally. And now there are also reports from late April 2026 that they're building an actual smartphone, which is a pretty big change from what they were saying before.
What Does "No Screen" Actually Mean?
So the first device, the small audio one codenamed "Sweetpea", is the one getting the most attention right now, and the plan is for it to have no screen or almost no screen.
This part confuses a lot of people. If there's no screen, then what exactly does it do?
The idea is simple. You talk to it. It listens, it responds, and it does things for you. You don't tap apps, you don't scroll through menus. You just ask. Say you want to book a cab, set a reminder, find out the weather, or send a message, instead of opening four different apps, you just tell the device and it sorts it out.
Think of it like having a really, really capable assistant that fits in your pocket. Except this assistant doesn't forget what you told it earlier and doesn't need you to spell everything out.
It also has cameras and microphones built in, so it can see and hear what's around you. The way it's been described is that it's "contextually aware", meaning it pays attention to your surroundings, not just your words. Say you're at a restaurant and you hold it up toward a dish, it can tell you what's in it. Or you're at a meeting, it's listening, and it can help you remember what was discussed later. There's something a little uncomfortable about that last bit, which I'll get to.
And Now There's a Smartphone Too?
This is where it gets a bit confusing. For a while, everyone was saying OpenAI is NOT making a phone. They're making something different, something new. Jony Ive himself had talked about wanting to move away from smartphones because he thinks they've caused real problems, people addicted to their screens, constantly distracted, that sort of thing.
But then, around late April 2026, a well-known supply chain analyst named Ming-Chi Kuo, who has a pretty decent track record predicting Apple's product plans, said that OpenAI is also working on an actual smartphone. A phone with chips from Qualcomm and MediaTek, manufactured by a company called Luxshare, with mass production possibly starting in the first half of 2027.
And a week after that, there were more reports saying OpenAI is actually fast-tracking this phone. So yeah, the plan keeps changing.
The phone, if it exists, would be different from your current Android or iPhone in one main way: instead of apps, it would use what are called "AI agents."
What Is an AI Agent? (This Part Is Important)
Okay, forget the hardware for a second. The actual big idea here is about how we interact with our devices.
Right now, your phone is basically a collection of apps. Want to order food? Open Uber eats. Want to book a flight? Open MakeMyTrip. Want to pay someone? Open Apple pay. Each app is its own little world. You have to open it, log in, navigate through it, do the thing, come back.
An AI agent is different. It's a piece of software that can do tasks on your behalf across multiple services, without you having to open anything. So instead of opening Uber eats, you'd just say "order me pizza from that place we ordered from last Tuesday", and the agent figures out the restaurant, places the order, handles the payment, and tells you when it's coming.
Sounds good on paper. And honestly it does make things easier, at least for people who find apps annoying to navigate.
The problem is that for an agent to do all this, it needs access to your accounts, your location, your payment information, your messages. A lot of access. And that's where things get tricky.
Is This Actually Useful or Just Hype?
Okay, I'll be honest here, because there's a history of devices like this failing badly.
In 2024, a company called Humane launched something called the AI Pin, a small wearable device that you could clip to your chest and talk to. It was supposed to be this revolutionary thing that would replace your phone. It got terrible reviews. The AI was too slow, it couldn't do half the things it promised, and then in February 2025 HP bought Humane for a much smaller amount than it raised and basically shut everything down. The device got permanently bricked, meaning it just stopped working. People who paid $699 for it got nothing.
So there's reason to be cautious.
OpenAI's situation is different in one way, their AI is genuinely much better than what Humane had. ChatGPT actually works. People use it. That's not nothing. But building hardware is a whole other problem, and OpenAI has never shipped a physical product before. The report from late 2025 mentioned that one of their biggest struggles is server capacity, they're already straining to run ChatGPT for all their users, and adding a device that's "always on" and always listening would need way more computing power than they currently have.
The no-screen device sounds nice in theory but I'm still not 100% sure who it's for. It probably works great if you're someone who already uses voice commands a lot. But most people I know, including me, still prefer typing. And the always-listening part genuinely makes me uncomfortable. The idea that a device with a camera and microphone is with you all day, collecting your "full real-time state", your location, your conversations, your habits, that's a lot to hand over.
How Is This Different from My Phone or Laptop?
Let me just put this simply.
Your current phone has apps. You control everything manually. You decide what to open, what to do, when to close it. The phone doesn't really "know" what you're trying to accomplish, it just runs whatever you tap on.
OpenAI's vision is a device where you describe what you want to accomplish and it handles the steps. Less like a tool you operate, more like a colleague you give instructions to.
Your current laptop or phone also treats you as a "user", meaning it waits for you to do something. The OpenAI device would be "proactive," meaning it might notice things and bring them to your attention without you asking. Like if it notices you have a meeting in 20 minutes and traffic is bad, it could tell you to leave now.
Sounds helpful. Also sounds like something that could get annoying very fast if the AI guesses wrong about what you need.
Another difference: your phone runs apps made by thousands of companies. The OpenAI device, at least the no-screen one, would run almost entirely on OpenAI's own AI. So if OpenAI's AI is wrong or makes mistakes and it does, there's not really a backup.
And the smartphone they're apparently building, that one would still look like a regular phone but the main way of using it would be through agents instead of apps. So you'd still have a screen, but the experience would feel totally different from how you use a phone today.
When Is All of This Actually Coming?
This is where I have to say: nobody fully knows.
The no-screen device was supposed to come in the second half of 2026. Then there were reports of delays. Then OpenAI's global affairs chief said in January 2026 that it's "on track." A court filing mentioned the device won't ship before February 2027 at the earliest. Also, apparently there's a trademark dispute with a hearing aid startup called iYo, so they can't even call it "io" anymore.
The phone that Kuo is reporting about mass production is now being targeted for the first half of 2027, maybe. Specifications aren't even final yet. Full consumer availability could be 2027 or 2028, depending on who you believe.
So if you're thinking of waiting for this before buying your next phone, I'd say don't. Buy the phone. This is at least two years away, probably more, and the final product might look completely different from what's being reported right now.
Should Normal People Care About This?
Yes and no.
If you're comfortable with technology, this will probably be interesting to watch. The idea of AI agents doing tasks for you is coming regardless of whether OpenAI's device succeeds, Google, Apple, and Samsung are all working on similar things. Your next phone, two or three years from now, will almost certainly have some version of this agent idea built in.
If you're not a tech person, the honest answer is that the most useful version of this, where you just talk and things happen, is still being figured out. The companies building it haven't solved the basics yet. So right now, it's more of a "coming soon" situation than something you need to think about today.
But the underlying shift is real. The way we interact with technology is changing. Apps and menus and tapping around, that whole model is slowly getting replaced by something more conversational. Whether OpenAI's specific devices are the ones that make it happen, or whether it's Google or Apple or someone else, that's hard to say.
What I do think is that the no-screen idea is probably ahead of its time. People are not ready to give up their screens yet. But the agent idea, where AI helps you get things done without juggling ten apps, that's something people will actually use once it works reliably.
One More Thing: The Privacy Problem
A device that's always on, always listening, always watching, always tracking your location and your habits, that's an enormous amount of personal information going to one company. OpenAI is an American company. Their track record on privacy is... not great, honestly. They've had incidents where conversation data was not handled the way users expected.
This isn't me saying don't use it. I'll probably try it when it comes out. But going in with open eyes matters. Before handing over that much access to any device or company, it's worth asking: what exactly is being stored, where, and for how long?
That question doesn't have a clear answer yet, and OpenAI hasn't addressed it in any detail publicly. They said they're still figuring out how the device will handle data privacy, which is a strange thing to admit when you're also saying the device might launch next year.
So yeah. Interesting times. Just don't hold your breath waiting for the thing to actually show up on a shelf.