OpenAI Is Making a Smart Speaker With a Camera — And It Costs $6.5 Billion to Get Here

So let me get this straight. OpenAI — the company that still isn't profitable, that reportedly burns through billions of dollars a year, and that has never shipped a physical product in its life — is building a smart speaker with a camera in it. Designed by Jony Ive. Powered by ChatGPT. And according to Sam Altman, it's going to be "the coolest piece of technology that the world has ever seen." I'm gonna need you to sit with that quote for a second. The coolest piece of technology. That the world. Has ever seen. For a speaker.

Look, I try really hard not to be cynical about new tech. I genuinely want cool stuff to exist. But when The Information dropped this report yesterday detailing OpenAI's hardware plans — a smart speaker, smart glasses, and even a smart lamp — I had some questions. A lot of questions, actually. So let's go through all of this, because the story behind how we got here is honestly more interesting than the product itself.

The $6.5 Billion Backstory

To understand the speaker, you need to understand the deal that made it possible. In May 2025, OpenAI acquired Jony Ive's hardware startup io for $6.5 billion. Jony Ive — the guy who designed basically every iconic Apple product from the iMac to the iPhone to the AirPods. He left Apple in 2019, started io with Altman, and OpenAI bought the whole thing. That's $6.5 billion for a company that, at the time, had a prototype that nobody had seen and a vague promise of "a new generation of AI-powered computers."

Altman and Ive did a press tour where they described their device as "peaceful," "an active participant that's not annoying," and something that would "make people feel joy." They also specifically said they didn't want a screen on it. No screen. Just voice, cameras, and sensors. A "third core device" alongside your phone and laptop. At one point Altman called it a "lakeside cottage for your digital life." I'm not making that up.

The original plan was to ship something by end of 2026. That didn't happen. Court filings from a trademark dispute revealed the device has been pushed to February 2027 at the earliest. They've dropped the "io" name entirely after getting sued by a hearing aid company. And they haven't created any packaging or marketing materials yet. So we're about a year out from launch and they're still figuring out what to call it.

The Speaker: What We Actually Know

According to The Information's report, the first product OpenAI will ship is a smart speaker. Not the screenless pocket device that was originally teased, not the smart glasses, not the lamp — a speaker. And here's what we know about it:

It'll have a built-in camera. Not just for looking pretty on your shelf — the camera is meant to identify objects near it, understand your surroundings, and support a facial recognition system similar to Face ID for authenticating purchases. It'll also have a mic, obviously, for talking to ChatGPT. But here's the kicker: reportedly, there's no wake word. The speaker is designed to be always aware of its surroundings. Always listening. Always watching. It picks up on context from your conversations and your environment without you explicitly asking it to.

In an internal presentation, OpenAI employees were told the speaker would observe users and suggest actions to help them achieve goals. The example they gave? If you have a morning meeting, the speaker might suggest you go to bed early. Which is... I mean, I don't know about you, but I don't need a $300 camera speaker to tell me to go to sleep. My mom does that for free.

The price target is $200 to $300, which puts it right in Amazon Echo Show and Google Nest Hub territory. Except those devices have screens, established ecosystems, years of smart home integrations, and don't have cameras pointed at your living room without a wake word. So OpenAI is going to need to offer something dramatically better to justify the comparison.

The Jony Ive Factor: Genius or Problem?

I wanna talk about Jony Ive's role here because it's more complicated than "legendary designer makes cool thing." Ive's design firm LoveFrom is technically separate from OpenAI — LoveFrom provides hardware designs, and OpenAI's engineers have to actually build them. According to The Information, OpenAI employees have complained about LoveFrom's secrecy and slow pace on design revisions. Ive apparently makes the final call on almost all design choices.

Former Apple designer Evans Hankey is leading industrial design. Other ex-Apple people on the team include Tang Tan and Scott Cannon. Eddy Cue's son Adam is working on OpenAI software. They've also partnered with Luxshare — one of Apple's key manufacturing partners — to produce early units. If you squint, this whole thing looks like an Apple product that Apple didn't make.

But here's the thing people forget about Jony Ive: his best work happened under Steve Jobs, who was famously great at telling Ive "no." Jobs constrained Ive's wilder ideas and forced practical decisions. Without that counterweight, Ive's track record is... mixed. The later MacBook Pro keyboards that everyone hated? That was Ive prioritizing thinness over usability. The Mac Pro trash can that couldn't be upgraded? Ive. Some of his unconstrained design choices at Apple were his worst ones.

And now his design partner is Sam Altman, who is not exactly known for telling people "no" to ambitious ideas. If anything, Altman is the opposite of Jobs in that regard — he's the guy who calls a speaker prototype "the coolest piece of technology the world has ever seen." That's not the energy of someone who's going to rein in over-design. That's the energy of someone who's going to let the design team do whatever they want and then hype it to the moon.

The Technical Problems Nobody Wants to Talk About

The Financial Times reported back in October that the OpenAI-Ive partnership was hitting serious technical walls, and the problems are pretty fundamental:

Compute. Running ChatGPT-level AI on a consumer device that costs $200 is a massive challenge. Amazon and Google have spent years and billions of dollars building the cloud infrastructure to keep Alexa and Google Home responsive. OpenAI is already straining to handle ChatGPT's existing demand. Adding millions of always-on smart speakers to that load? Someone close to Ive told the FT that "compute is another huge factor for the delay." Yeah, I bet it is.

Privacy. An always-on camera with no wake word in your home. I don't care how much you trust OpenAI — that's a tough sell. Amazon took years of backlash over Alexa's always-listening mic, and that doesn't even have a camera. The always-on approach reportedly caused internal debates at OpenAI because the device would sometimes respond when it shouldn't. Getting AI to understand when you're talking to it versus just having a conversation with your partner is still an unsolved problem.

Personality. This sounds weird but it's apparently a real issue. How should an AI speaker "behave"? When should it speak up and when should it stay quiet? Should it be proactive or reactive? These are basic interaction design questions that OpenAI reportedly still doesn't have answers to. Amazon spent years teaching Alexa basic etiquette and it still gets it wrong constantly. OpenAI is trying to do something way more ambitious with way less experience.

The AI Hardware Graveyard

I can't write this article without talking about all the AI hardware products that have tried this exact same thing and failed spectacularly. Because the graveyard is getting pretty full.

Every single one of these products was backed by massive companies with enormous resources, and every single one failed. The common thread? They all asked consumers to put a camera and/or always-on microphone in their home connected to a company's cloud, and consumers said no thanks. OpenAI is now asking for both at once, from a company with no hardware track record, running on cloud infrastructure they don't fully control.

Now — is OpenAI's AI better than what Amazon or Facebook had? Absolutely. ChatGPT is leagues ahead of old Alexa. But "the AI is good" has never been the problem. The problem is trust, privacy, and whether the product actually does something your phone can't already do. None of these dead products failed because the AI wasn't smart enough. They failed because people didn't want them in their lives.

The Full Hardware Roadmap

The speaker is just the first thing. Here's what OpenAI has planned according to the reporting:

The Competition Is Already Here

Here's the thing — by the time OpenAI ships this speaker in early 2027, the smart speaker market will have already leveled up without them. Google's new Home Speaker with Gemini built in is shipping in a couple months. Amazon has Alexa Plus, the AI-powered subscription tier. Apple's got the HomePod with Siri improvements coming. All of these companies have years of smart home integrations, massive installed bases, and the cloud infrastructure to handle it.

What does OpenAI have? ChatGPT. That's the whole value proposition. ChatGPT is really, really good at conversation. It's great at answering questions, helping you think through problems, and generating content. But is "a better conversationalist" enough to get someone to ditch their Echo or Google Home? I'm genuinely not sure. Especially when Google's speaker already runs Gemini — which is very good — and Amazon is integrating Claude into Alexa Plus.

And here's something Gizmodo pointed out that I think is worth highlighting: almost everything OpenAI described — looking at objects on a table, making purchases by voice, proactive suggestions — Amazon's Echo Show basically does already. The camera stuff with facial recognition is new, but is that a feature people are asking for? Or is it a feature that makes people uncomfortable?

My Take: Ambitious, But I Have Concerns

I'm gonna be real with you — I'm not writing this off. You'd be crazy to bet against Jony Ive's design skills combined with ChatGPT's intelligence. If anyone can make AI hardware that people actually want to use, it's probably this team. The $6.5 billion buys you a lot of talent and a lot of runway.

But I have three big concerns:

One: the camera with no wake word is going to be a massive PR problem. I don't care how good the product is — the "OpenAI is watching you in your living room" headlines write themselves. And after every privacy scandal of the last decade, consumers are not in a trusting mood.

Two: the financial situation. OpenAI is burning cash at a staggering rate and this speaker needs to not just exist — it needs to sell millions of units to justify that $6.5 billion acquisition and the 200+ person team. That's a really high bar for a first-generation product from a company with zero hardware experience.

Three: Altman's hype is getting ahead of the product. Calling it "the coolest piece of technology the world has ever seen" before it's even launched is exactly the kind of thing that sets up for massive disappointment. Underpromise, overdeliver — that's how you win in hardware. Altman is doing the opposite.

That said, I've seen enough tech cycles to know that the product everyone laughs at in the rumor phase sometimes turns into the product everyone owns three years later. The first iPhone got mocked. The first AirPods got mocked. If OpenAI and Ive nail the execution and the product actually feels magical to use, none of my concerns will matter. But that's a very big "if."